REGISTER  |  SIGN IN
United States Patent Application20010026591
Kind CodeA1
Keren, Avishai ; et al.October 4, 2001
Multimedia stream compression
Abstract
A computer may be remotely accessed. At a first location, display commands are generated. The display commands are covered into a compressed video data stream. Each display element (5) is checked if it is encoded (52). If object is encoded, it is transcribed into MPEG (54). The image is adjusted for display (56) and compression (58). Additional steps of motion determination (60), change detection (62), compression depth and frame determination (66, 68) are executed. Then the data is transmitted to a second location. The display commands are decompressed and displayed as an image at the second location.
Inventors:Keren; Avishai (Rosh Ha'ayin, IL), Feder; Meir  (Herzelia, IL), Paz; Ofir  (Rishon Lezion, IL)
Correspondence Name and Address:WORKMAN, NYDEGGER & SEELEY 1000 Eagle Gate Tower 60 East South Temple
RICK D. NYDEGGER
Salt Lake City
UT
84111
US
Series Code:770767
Filed:January 25, 2001
U.S. Current Class:375/240.26
U.S. Class at Publication:375/240.26
Intern'l Class:H04N 007/12
Claims

What is claimed and desired to be secured by United States Letters Patent is:
1. A method of generating a compressed video stream, comprising: receiving a plurality of display commands; and directly converting said commands into a compressed video stream, without first generating a display raster.

2. A method of generating a compressed video stream, comprising: receiving a plurality of display commands; and directly converting said commands, into said compressed video stream, wherein said converting comprises motion estimation and wherein said motion estimation is performed directly on said commands without first generating a display raster.

3. A method of generating a compressed video stream, comprising: receiving a plurality of display commands; and directly converting said commands into said compressed video stream, wherein said converting comprises change detection and wherein said change detection is performed directly on said commands without first generating a display raster.

4. A method of motion estimation for image stream compression, comprising: providing a plurality of display commands; identifying display manipulation commands among said commands which manipulation commands manipulate image portions in said image stream, said portions being generated by other display commands among said plurality of display commands; and generating at least one motion vector, responsive to said identified display manipulation commands.

5. A method according to claim 4, wherein said display manipulation commands include a scroll command.

6. A method according to claim 4 or claim 5, wherein said display manipulation command is generated as a direct result of a display of a downloaded image in a WWW browser.

7. A method according to any of claims 4-5, comprising reducing a difficulty of said motion vector generation by modifying said display manipulation commands.

8. A method according to claim 7, wherein scroll-type display modification commands are limited to multiples of a compression block size dimension.

9. A method of change detection for image stream compression, comprising: providing a plurality of display commands; identifying commands of said plurality of commands which affect image portions in said image stream; identifying said image portions; and determining if a change in one of said image portions requires image information to be included in said image stream.

10. A method according to claim 9 wherein said determining comprises determining if said change is greater than a change threshold.

11. A method of generating a compressed video stream, comprising: providing a plurality of display commands; directly setting values in a transform space, responsive to said display commands; and creating a compressed video stream utilizing said set values.

12. A method according to claim 11, wherein said transform space comprises a DCT space.

13. A method according to claim 11, wherein directly setting comprises utilizing transform coefficients from a look-up table.

14. A method according to claim 11, wherein directly setting comprises copying transform coefficients from a look-up table of transformed graphical primitives.

15. A method according to claim 11, wherein directly setting comprises copying transform coefficients from a cache of transformed image portions.

16. A method according to claim 11, wherein directly setting comprises rendering said display command into said transform space utilizing an analytical formula associating a graphical primitive defined by said display command with said transform space.

17. A method according to claim 11, comprising, modifying at least one of said display commands so that a graphical primitive defined by said at least one display command does not straddle an image block, for which block a set of transform coefficients are defined according to a compression protocol utilized by said compressed image stream.

18. A method of generating a catalog, comprising: providing a plurality of display elements; generating a compressed video representation of each one of said elements; and storing said compressed representations in a memory, indexed by element.

19. A method according to claim 18, wherein said display elements comprise GUI (graphic user interface) objects.

20. A method of generating a catalog, comprising: receiving, during interaction with an Internet a plurality of objects to be displayed; generating a compressed video representation of each one of said objects; and storing said compressed representations in a cache of objects, whereby when one of said objects is received again from said Internet at a later time, a earlier stored compressed representation is used for display generation.

21. A method according to claim 20, wherein said stored objects comprise at least one JPEG image.

22. A method according to claim 20, wherein said stored objects comprise at least one uncompressed image.

23. A method according to claim 20, wherein said stored objects comprise a WWW page.

24. A method of generating a compressed video stream, comprising: generating a plurality of display commands which represents a display, by a program; modifying at least one of said display commands, independently of said program; and generating a compressed video stream utilizing said display commands.

25. A method according to claim 24, wherein said modifying comprises modifying said commands to responsive to limitations of a display device on which said compressed video stream is to be displayed.

26. A method according to claim 24, wherein modifying comprises modifying said commands to reduce bandwidth requirements of said compressed video stream.

27. A method according to claim 24, wherein modifying comprises modifying said commands to reduce resource requirements for compressing said compressed video stream.

28. A method according to any of claims 24-27, wherein at least one of said display commands represents an object and wherein said modifying comprises changing said at least one display command such that the object is moved relative to its original display position.

29. A method according to claim 28, wherein said compressed video stream comprises utilizes blocks having boundaries and wherein moving comprises moving said object to match at least one compression block boundary.

30. A method according to any of claims 24-27, wherein at least one of said display commands represents an object and wherein said modifying comprises replacing said object with a different object.

31. A method according to claim 30, wherein replacing said object comprises replacing said object with a compressed representation of said different object.

32. A method according to claim 30, wherein said object comprises a text object.

33. A method according to claim 30, wherein said object comprises a background of said display.

34. A method according to claim 30, comprising analyzing said object to determine a closest suitable replacement object.

35. A method according to any of claims 24-27, wherein modifying said object comprises changing a font definition for said object.

36. A method according to any of claims 24-27, wherein modifying said object comprises modifying at least one color of said object.

37. A method according to claim 36, wherein modifying a color comprises reducing a spatial resolution of said colors.

38. A method according to claim 36, wherein modifying a color comprises reducing a color range resolution of said colors.

39. A method according to any of claims 24-27, wherein modifying said object comprises reducing a spatial resolution of said object.

40. A method according to any of claims 24-27, wherein at least one of said display commands comprises a scrolling command and wherein said modifying comprises increasing a granularity of said scrolling.

41. A method according to claim 40, wherein said increasing a granularity comprises limiting said scrolling command to multiples of compression blocks size of said compressed stream.

42. A method of generating a compressed video stream, comprising: generating a plurality of display commands which represents a display; generating a plurality of transform coefficients from said display commands, wherein said transform coefficients are quantized and wherein said quantization is modified responsive to limitations a display device on which said compressed video stream is to be displayed; and creating a compressed video stream utilizing said coefficients.

43. A method according to claim 42, wherein said coefficients are quantized at a lower resolution in a vertical direction of said display.

44. A method of generating a compressed video stream, comprising: providing a display definition including a compressed object; and converting said display definition into a compressed video stream, wherein converting includes transcribing at least a portion of said compressed object, without decompressing said compressed object.

45. A method according to claim 44, wherein transcribing comprises reducing a quality level of said compressed object.

46. A method according to claim 44 or claim 45, wherein said compressed stream and said compressed object are compressed utilizing a same transform and wherein transcribing comprises copying transform coefficients from said compressed object to said compressed video stream.

47. A method according to claim 44 or claim 45, wherein said compressed stream and said compressed object are compressed utilizing a motion estimation representation and wherein transcribing comprises copying motion estimation vectors from said compressed object to said compressed video stream.

48. A method according to claim 44 or claim 45, wherein converting said display definition comprises directly converting said display representation into said compressed video stream, without generating an intermediate image raster.

49. A method according to claim 44 or claim 45, wherein said compressed object comprises a second compressed video stream, covering only a portion of said display.

50. A method according to claim 44 or claim 45, wherein said compressed object comprises a compressed image, covering only a portion of said display.

51. A method according to claim 44 or claim 45, wherein said compressed object comprises a compressed audio stream.

52. A method of image processing, comprising: providing an image to be compressed for transmission; simultaneously compressing and processing said image; transmitting said image to a remote location; and decompressing said processed image at said remote location.

53. A method according to claim 52, wherein said simultaneously compressing and processing comprises processing said image by modifying a quantization of at least one transform coefficient of said image.

54. A method according to claim 52, wherein said modifying comprises increasing a quantization granularity for vertical coefficients.

55. A method of generating a compressed video stream, comprising: accumulating a plurality of display commands; identifying at least one display command of said plurality, whose effect is mooted by a later accumulated display command; and converting said commands into said compressed video stream, ignoring the mooted display command.

56. A method according to claim 55, wherein said mooted display command is ignored if its effect on said compressed video stream, in view of said later command, is below a threshold value.

57. A method according to claim 55, wherein converting said commands comprises combining display commands.

58. A method according to any of claims 55-57, wherein said converting comprises directly converting said commands into said compressed video stream without first generating an image raster.

59. A method of video compression, comprising: generating a display, which generation is refreshed at a generation refresh frame rate; compressing, transmitting and decompressing said display; and displaying said display at a second refresh rate other than said generation refresh rate and wherein said refresh is not synchronized to said display refresh rate.

60. A method according to claim 59, wherein said compressing is un-synchronized with said generation refresh frame rate.

61. A method according to claim 59, wherein said transmitting is un-synchronized with said compressing.

62. A method of generating a compressed video stream, comprising: generating a plurality of display commands; accumulating at least one display command; generating a compressed video stream from said accumulated display commands, wherein accumulating comprises accumulating a number of display commands responsive to instantaneous available resources.

63. A method of generating a compressed video stream, comprising: generating a plurality of display commands; accumulating at least one display command; generating a compressed video stream from said accumulated display commands, wherein accumulating comprises accumulating a number of display commands responsive to a desired output frame rate.

64. A method according to claim 62 or claim 63, wherein generating a compressed video stream is un-synchronized relative to said generation of display commands.

65. A method according to claim 64, wherein said display commands are generated in sets, each set associated with a display frame and wherein said accumulating, in a single accumulation set, display commands originating from different sets, wherein said single accumulation set is utilized for generating a single frame of said compressed video stream.

66. A method according to claim 62 or claim 63, comprising assigning a priority to a display command.

67. A method according to claim 66, comprising reordering said accumulated display commands, responsive to said priority, wherein generating a compressed video stream comprises generating a first frame in said compressed video stream utilizing an originally later display command, prior to generating a frame in said stream utilizing an originally earlier display command.

68. A method of generating a compressed video stream, comprising: generating a plurality of display commands; generating a compressed image portion responsive to said display commands; and assembling a compressed video stream from said compressed image portion and from at least one previously compressed image portion, created prior to generation of said display commands.

69. A method according to claim 68, wherein said previously compressed image portion comprises an image portion created for a previous image of said video stream.

70. A method of asynchronous video stream compression, comprising: providing a first data source and a second data source; converting said first data source into a first portion of a compressed video stream; converting said second data source into a second portion of a compressed video stream; and mixing said first portion and said second portions to create a single compressed video stream, wherein said first portion and said second portion temporally overlap in said single compressed video stream.

71. A method according to claim 70, wherein said first portion and said second portion utilize different frame rates.

72. A method according to claim 70, wherein said first portion and said second portion occupy different spatial areas in a display defined by said single compressed video stream.

73. A method according to any of claims 70-72, wherein said first and said second data sources comprise sets of display commands.

74. A method according to claim 73, wherein converting said first data source and converting said second data source comprises directly converting said data sources without generating an intermediate image raster.

75. A method according to any of claims 70-72, wherein at least one of said first data source and said second data source comprises an image source.

76. A method according to any of claims 70-72, wherein said first data source comprises a display portion affected by a command entered by a user who views said single compressed video source.

77. A method according to claim 76, comprising identifying portions of a data display which are affected by said user command.

78. A method of generating a compressed video stream, comprising: executing a program, which program generates a set of display commands, representative of a display; identifying changes in said display which are responsive to at least one type of user command; and converting said display commands into a compressed video stream, wherein said changes are inserted into said compressed video stream at a higher frame rate than other changes in said display.

79. A method according to claim 78, wherein said type of user command comprises a pointing device command.

80. A method according to claim 78, wherein said changes comprises an indication of a selection of a GUI (graphical user interface) element.

81. A method according to any of claims 78-80, comprising analyzing said user command to determine display commands which effect said identified changes.

82. A method of display control, comprising: transmitting a request, from a generator of a compressed video stream, to a source of display commands, which requests relates to said compression; and modifying a generation of display commands, at said source, responsive to said request.

83. A method according to claim 82, wherein said source of display commands comprises a WWW browser.
Description


RELATED APPLICATIONS

[0001] This application claims priority to under 35 USC .sctn.120 or .sctn.365(c) of PCT application No. PCT/IL98/00349 entitled Remote Computer Access filed Jul. 27, 1998.

BACKGROUND OF THE INVENTION

[0002] 1. The Field of the Invention

[0003] The present invention relates to remote operation of computer systems and in particular to remote access to a computer running an Internet browser.

[0004] 2. The Relevant Technology

[0005] Various methods of remotely accessing a computing server have been used in the past. One of the first methods utilized a server and terminals. Input is entered at the terminals, transmitted to the server and the results are send back over data lines to the terminals. The terminals can be "dumb" terminals, such as a VT52, with no processing power except to display the data, or they can be smart terminals, such as an IBM 3270, which can do some basic form filling without requiring intervention of the server.

[0006] As computing power has grown cheaper, various other strategies have been used. In the X11 graphical networking protocol, a display server sends bitmaps to a display client and receives input events from the client In the NeWS graphical networking protocol, a server sends short code segments to a client computer. The code segments perform display and input handling functions at the client and, when needed, send input to and receive display data from the server.

[0007] A somewhat different type of remote computer access is Internet-TV. A television set is upgraded using an electronic box so that the television may be used to access the Internet. In one system, Internet related data is transmitted to and from the electronic box using protocols which transmit data over a cable network.about. In some instances, the data is transmitted using a data channel provided by an MPEG compression protocol. In one system, the box is a computer, which performs the Internet communication via the cable network and which uses the television as a display. In another system, a server at a cable center, sends graphical display commands to the box and the box generates a video signal responsive to the commands. Typically, the display commands are a subset of Java or HTML. In another system, the data comprises Java scripts which are executed by the box.

[0008] Cable networks have also long been used to send video and/or audio programs on demand to a particular subscriber. The video programs are compressed before sending so that a wider bandwidth is available over the cable networks. A set-top box at each subscriber is hardwired (and/or programmed) to decompress the video program and convert it into a video signal for the television. Such video on demand may be a live performance or may be retrieved from an archive. It is also possible to view video segments via the Internet.

[0009] Generally, when video is transmitted, a same bandwidth is allocated to a plurality of video channels. In packet based video transmission, subscribers and/or channels that require a higher data rate receive more packets per second than subscribers who need a lower data rate. A company named "iMedia" uses a method of statistical multiplexing in which bandwidth is dynamically allocated between video channels, based on signal characteristics of the video being transmitted. The allocation does not depend on the content per-se of the video channels. In fact, the statistical multiplexing is preferably affected without decompressing the images.

[0010] Related patent publications include U.S. Pat. Nos. 5,612,742, 5,687,257 and 5,115,309 and PCT publication WO 97/12486, the disclosures of which are incorporated herein by reference.

[0011] The following articles, "Polygon Assisted JPEG and MPEG Compression of Synthetic Images", by Marc Levoy, Proceeding of Siggraph 1995, "Model Based Motion Estimation for Synthetic Animations", by Maneesh Agrawala, Andrew C. Beers and Navin Chaddha, in ACM Multimedia 1995, and "Compression Performance of the Xremote Protocol", by John Danskin and Pat Hanrahan, Proceedings 1994 Data Compression Conferences, the disclosures of which are incorporated herein by reference, describe methods compressing animation into MPEG format, on a single machine. In these methods the compression is faster since the "video source", the animation, is completely known.

BRIEF SUMMARY OF THE INVENTION

[0012] One object of some embodiments of the present invention is to provide a client server protocol which operates via a cable network. Preferable, the server is used to connect to the Internet. Preferably, the client comprises a television and a set-top box.

[0013] An object of some preferred embodiments of the invention is to provide methods of more efficiently compressing video, which video is generated by a computer.

[0014] An object of some preferred embodiments of the invention is to provide methods of bandwidth allocation for cable-network systems where the data transmitted relates to computer communications. Preferably, the computer communications comprise compressed screen displays.

[0015] One aspect of some preferred embodiments of the present invention relates to providing remote computer access using compressed video. Preferably, one or more clients are connected to a single server using a cable network. Preferably, displays are sent from the server to the clients using compressed video, preferably using an MPEG II compression protocol. Preferably, the clients provide feedback to the server using data transmission over the cable network. Alternatively, the feedback may be provided over a telephone line.

[0016] Another aspect of some preferred embodiments of the invention relates to efficient video compression, especially MPEG II compression, of computer generated displays. Efficient compression may be faster, of a higher quality and/or be obtained using a cheaper hardware and/or software base. In one preferred embodiment of the invention, such efficient compression is achieved by bypassing motion estimation algorithms. Such bypassing may be achieved by taking note of display manipulation commands and tracking which portions of a display are changed and which are shifted. Display manipulation commands include, for example, scrolling commands and pan commands.

[0017] Alternatively or additionally, efficient compression is achieved by maintaining the compression of video, audio and/or still images which are retrieved by the server and sent to the client, instead of decompressing them and then compressing them again prior to sending them.

[0018] Alternatively or additionally, efficient compression is achieved by modifying a display, which would have been used for a user working on the server, so that it can be more efficiently compressed.

[0019] Alternatively or additionally, efficient compression is achieved by direct drawing of graphical elements of a display into a DCT space, instead of- or in addition to- drawing into an image space and then performing a transform.

[0020] Alternatively or additionally, efficient compression is achieved by reusing portions of a display and/or a DCT space, between images and/or between clients.

[0021] Another aspect of some preferred embodiments of the invention relates to image processing an image and/or a video stream. In a preferred embodiment of the invention, image processing of an image is combined with a compression stage by providing an image, compressing the image and image processing the image by manipulation of compression parameters and/or by processing of the compressed data. It should be appreciated that it would have been possible to process the image before compressing. However, in some cases, the image processing is more computational efficient by processing the compressed data. In a preferred embodiment of the invention, an image is low-pass filtered by reducing the quantization rate of low frequency components. In a preferred embodiment of the invention, only the vertical frequency components are modified, to reduce flicker in a reconstructed image. In another example a different wavelet resolution is used in a vertical direction of an image.

[0022] Another aspect of some preferred embodiments of the invention relates to asynchronic image compression and transmission. In a preferred embodiment of the invention, an image or an image portion is compressed in an asynchronic manner, for example, depending on available bandwidth, available computational power, indication of changes in the image, desirable image quality and/or type of change in the image. In a preferred embodiment of the invention, the display rate and/or the image generate rate are unaffected by the asynchronic characteristics of the data compression.

[0023] In a preferred embodiment of the invention, a compression step is prompted by an event which marks a change in the image. In a preferred embodiment of the invention, only the portions of the image which are affected by the event are compressed. A new image frame is created and sent to a decompression and display unit. Such a new image frame may comprise a completely new frame or a frame which defines changes over one or more previous frames. In a preferred embodiment of the invention, two change events may be combined in a single new frame. In some cases, one change event may moot a previous change event, so that only the changes associate with one event are processed.

[0024] In a preferred embodiment of the invention, the system may operate at several different frame rates simultaneously: image generation frame rate, actual compression frame rate, maximum compression frame rate and display frame rate. In a preferred embodiment of the invention, each of these frame rates may be independently varied, for example, the image generation rate may depend on a computer program which generates the image. The actual compression rate may depend on available computing power and/or on whether there were any changes in the frame. The maximum compression rate may be the maximum frame rate which is allocated for a particular user and may depend on a user request, available bandwidth and/or payment schedule. The display rate may depend on the characteristics of the display device, be it NTSC, PAL or HDTV. In a preferred embodiment of the invention, the actual compression rate and the display rate are synchronized by buffering the actually sent frames with "no change" frames, to maintain the display frame rate constant. In a preferred embodiment of the invention, the maximum frame rate determines the number of frames sent, with the balance between the maximum frame rate and the compression frame rate also being buffered using "no change" frames.

[0025] Alternatively or additionally, the MPEG decompression module may be modified to work with variable frame rates.

[0026] Another aspect of some preferred embodiments of the invention relates to bandwidth allocation between a plurality of clients. In accordance with one preferred embodiment of the invention, the compression depth is traded-off between the clients. Preferably, each client is provided with a minimum bandwidth according to his needs.

[0027] Alternatively or additionally, bandwidth is allocated by providing each client with a different frame rate. Typically, a 25/30 Hz (changed) frame rate is not needed in any computer application, so that a lower frame rate may generally be used for such applications.

[0028] Alternatively or additionally the bandwidth is allocated responsive to the amount of motion in the display for each client.

[0029] Another aspect of some preferred embodiments of the invention relates to adjusting the displayed information to match a particular compression required and/or display characteristics of a television used to display the data. In one example, a cursor flashing rate may be changed so that a lower frame rate is required. In another example, the colors of a displayed portion are adjusted so that compression of color components is more efficient. In still another example, display elements, such as bullets and letters are moved so that they do not straddle compression-block boundaries. In another example, a display background may be replaced by a different display background, for example to simplify compression or to utilize a pre-compressed background. Preferably, the background is compared to a set of stored background portions to determine which background is most similar. In a preferred embodiment of the invention, if all the pre-stored backgrounds are different by more than a threshold amount, the original background may be used. Alternatively or additionally, the original background may be modified to simplify compression. In a preferred embodiment of the invention, a background is analyzed to determine if it is unique and should not be replaced. Such an analysis may include, inter alia, analyzing the size of the background image, analyzing the contrast and/or analyzing the colors of the image. Generally, a garish, large and/or high contrast background is a unique background. Plain, simple and low-contrast backgrounds are generally "standard" and replaceable backgrounds. Preferably, when determining which background to replace the original background with, a plurality of parameters of the original background are used to select a match. Such parameters can include main color (a replacement background can be set to that color), spatial frequency spectrum (a background with a similar spectrum can be selected) and size (a background with a similar size and/or aspect ration may be selected). Alternatively or additionally, other display elements, such as icons, buttons, menus and fonts may also be replaced and/or modified in similar ways to a background.

[0030] Another aspect of some preferred embodiments of the invention relates to generating a plurality of content-unrelated compressed video streams from a single computer. In some preferred embodiments of the invention, such a single computer may include a plurality of processing units, sharing resources, such as memory and/or a main data bus. In a preferred embodiment of the invention, when a resource is not needed for the generation of one channel, that resource may be used by a second channel. In a preferred embodiment of the invention, at least some of the processing associated with the generation of both channels is performed on a single CPU. Preferably, a single computer generates displays for 4, 10 or even 20 channels, i.e., clients.

[0031] Another aspect of some preferred embodiments of the invention relates to a virtual device driver for an operating system, so that a plurality of full size windows may be displayed simultaneously on the virtual device, by a host computer. Preferably, each one of the full-size displays is converted into a separate and content-unrelated video stream. In a preferred embodiment or the invention, each window runs an instance of an Internet browser. In a preferred embodiment of the invention, the border and/or other portions of the browser are not shown. Alternatively or additionally, a control area may be optionally provided by server 16. Preferably, such a control area is overlaid on the display.

[0032] Another aspect of some preferred embodiments of the invention relates to broadcasting compressed video data packets representing a plurality of display channels, where some packets may be used for more than one channel. In one example, a packet including data corresponding to a menu portion of a web browser is utilized and decompressed by all the viewers which view that particular browser. In another example, when a plurality of viewers view a same TV or WWW channel, a temporal and/or spatial advertisement portion of the channel may be personalized for a particular viewer or for a group of users. The decompression unit at each different user or group or users preferably accepts different packets and thus may show different advertisements, on the same channel.

[0033] Another aspect of some preferred embodiments of the invention relates to treating WWW sites, radio stations, video on demand, cable channels and TV channels as different instances of multi-media presentation channels. In a preferred embodiment of the invention, some WWW sites, such as CNN, Microsoft and the local weather are set up as permanent channels. In some preferred embodiments of the invention the interactivity of a channel may be reduced in order to allow more channels in a given bandwidth.

[0034] In a preferred embodiment of the invention, such channels may be switched between and viewed on a single display device, such as a TV or a computer display. Alternatively or additionally, links may be defined between portions of such channels and portions of other such channels. Preferably, a user may follow such links to switch between channels. Alternatively or additionally, a user may view a plurality of such channels simultaneously and/or in an overlay. In a preferred embodiment of the invention, linking from a TV channel is enabled by an overlay which indicates which portions of the channel include links and/or where such links point. In a preferred embodiment of the invention, such links may be dynamic and/or track portions of the image. An audio channel or a TV channel may include an associated visual display of links and/or may include voice annotations informing of links. In a preferred embodiment of the invention, a link may be selected by a voice command. In a preferred embodiment of the invention, the links are personalized to match a particular viewer.

[0035] In a preferred embodiment of the invention, a web site may be used to browse TV channels. When a particular link on the web site is followed, a corresponding TV channel may be displayed. In a preferred embodiment of the invention, a cable operator may provide a WWW interface to browsing a TV channel catalog and especially a video- on-demand catalog.

[0036] In a preferred embodiment of the invention, advertisements in the multimedia channels are treated in aggregate. Thus, advertisements may be personalized for a particular viewer, depending on the user, for all the different types of multimedia channel. Generation of advertisements, tracking of user viewing and selection of advertisements shown may be effected using a single system, albeit possibly with different accounting for the different types of channels.

[0037] There is therefore provided in accordance with a preferred embodiment of the invention, a method of remote computer access, comprising:

[0038] executing a program at a first location, to generate display commands;

[0039] converting said display commands directly into a compressed video data stream;

[0040] transmitting said compressed data stream to a second location, remote from said first location;

[0041] decompressing said compressed data stream at the second location; and

[0042] displaying the decompressed data stream as an image at the second location.

[0043] Preferably, executing a program at a first location comprises executing a program on a general purpose computer at said first location. Alternatively or additionally, displaying the decompressed data comprises displaying the decompressed data on a TV set. Alternatively or additionally, transmitting said compressed data comprises transmitting said compressed data over a television distribution network. Preferably, said network comprises a cable network.

[0044] Alternatively or additionally, said decompressing comprises decompressing using a cable TV set-top box. Alternatively or additionally, said program comprises a word processor.

[0045] In a preferred embodiment of the invention, said program comprises a game program. Preferably, said program communicates with at least a second program executed on said general purpose computer.

[0046] In a preferred embodiment of the invention, said program accesses data stored in association with said general purpose computer. Preferably, said decompressed data stream comprises audio recordings. Preferably, the method comprises tracking access to said audio recordings for royalty payment assessment.

[0047] In a preferred embodiment of the invention, said program comprises a browser which accesses a third location, remote from said first and said second locations. Preferably, second remote location is accessed via an Internet.

[0048] In a preferred embodiment of the invention, said compressed video stream represents an entire TV display.

[0049] Preferably, the method comprises transmitting user inputs from said remote location to said first location, responsive to said display.

[0050] There is also provided in accordance with a preferred embodiment of the invention, a remote computing server system, comprising:

[0051] a server, executing a plurality of programs, each of which generates a set of display commands; and

[0052] a video compressor which receives the plurality of sets of display commands and generates a compressed video stream from each one of said sets. Preferably, the server comprises a mixing box which multiplexes said video streams unto a cable transmission network. Alternatively or additionally, said server comprises a mixing box which multiplexes said video streams unto a satellite transmission network.

[0053] There is also provided in accordance with a preferred embodiment of the invention, a method of video transmission, comprising:

[0054] executing, at a server computer, a plurality of programs, each of said programs generating a display responsive to an Internet connection; and

[0055] transmitting each of said displays to a different remote location, wherein said displays are transmitted as compressed video streams.

[0056] Preferably, each of said programs is connected to a different Internet address. Alternatively or additionally, each of said programs generates a set of display commands and said compressed video steams are directly generated from said sets of display commands. Preferably, the method comprises generating said compressed video streams responsive to known visual limitations at said remote locations. Alternatively or additionally, the method comprises generating said compressed video streams responsive to bandwidth limitations on said transmission. Alternatively or additionally, the method comprises degrading said display commands responsive to bandwidth limitations on said transmission.

[0057] There is also provided in accordance with a preferred embodiment of the invention, a multi-headed display generator, comprising:

[0058] at least one CPU running at least one program, each of said programs generating at least one set of display commands, wherein said programs generate in totality at least two sets of content independent display commands; and

[0059] at least one compressor which converts said two sets of display commands into two simultaneous compressed video streams,

[0060] wherein said compression of the said sets utilizes at least one shared resource of said generator.

[0061] Preferably, said resource comprises CPU resources. Alternatively or additionally, said resource comprises memory resources. Alternatively or additionally, said generator trades off the compression of one set of display commands with the compression of a second set of display commands.

[0062] Preferably, said trade-off comprises trading off quality between the two command sets. Alternatively or additionally, said trade-off comprises trading off frame rate between the two command sets.

[0063] In a preferred embodiment of the invention, said generator statistically multiplexes said compressed video streams onto a single transmission bandwidth.

[0064] There is also provided in accordance with a preferred embodiment of the invention, a method of generating a compressed video stream, comprising:

[0065] receiving a plurality of display commands; and

[0066] directly converting said commands into a compressed video stream, without first generating a display raster.

[0067] There is also provided in accordance with a preferred embodiment of the invention, a method of generating a compressed video stream, comprising:

[0068] receiving a plurality of display commands; and

[0069] directly converting said commands into said compressed video stream, wherein said converting comprises motion estimation and wherein said motion estimation is performed directly on said commands without first generating a display raster.

[0070] There is also provided in accordance with a preferred embodiment of the invention, a method of generating a compressed video stream, comprising:

[0071] receiving a plurality of display commands; and

[0072] directly converting said commands into said compressed video stream, wherein said converting comprises change detection and wherein said change detection is performed directly on said commands without first generating a display raster.

[0073] There is also provided in accordance with a preferred embodiment of the invention, a method of motion estimation for image stream compression, comprising:

[0074] providing a plurality of display commands;

[0075] identifying display manipulation commands among said commands which manipulation commands manipulate image portions in said image stream, said portions being generated by other display commands among said plurality of display commands; and

[0076] generating at least one motion vector, responsive to said identified display manipulation commands.

[0077] Preferably, said display manipulation commands include a scroll command. Alternatively or additionally, said display manipulation command is generated as a direct result of a display of a downloaded image in a WWW browser. Alternatively or additionally, the method comprises reducing a difficulty of said motion vector generation by modifying said display manipulation commands.

[0078] Preferably, scroll-type display modification commands are limited to multiples of a compression block size dimension.

[0079] There is also provided in accordance with a preferred embodiment of the invention, a method of change detection for image stream compression, comprising:

[0080] providing a plurality of display commands;

[0081] identifying commands of said plurality of commands which affect image portions in said image stream;

[0082] identifying said image portions; and

[0083] determining if a change in one of said image portions requires image information to be included in said image stream.

[0084] Preferably, said determining comprises determining if said change is greater than a change threshold.

[0085] There is also provided in accordance with a preferred embodiment of the invention, a method of generating a compressed video stream, comprising:

[0086] providing a plurality of display commands;

[0087] directly setting values in a transform space, responsive to said display commands; and

[0088] creating a compressed video stream utilizing said set values.

[0089] Preferably, said transform space comprises a DCT space. Alternatively or additionally, directly setting comprises utilizing transform coefficients from a look-up table. Alternatively or additionally, directly setting comprises copying transform coefficients from a look-up table of transformed graphical primitives. Alternatively or additionally, wherein directly setting comprises copying transform coefficients from a cache of transformed image portions. Alternatively or additionally, directly setting comprises rendering said display command into said transform space utilizing an analytical formula associating a graphical primitive defined by said display command with said transform space.

[0090] In a preferred embodiment of the invention, the method comprises modifying at least one of said display commands so that a graphical primitive defined by said at least one display command does not straddle an image block, for which block a set of transform coefficients are defined according to a compression protocol utilized by said compressed image stream.

[0091] There is also provided in accordance with a preferred embodiment of the invention, a method of generating a catalog, comprising:

[0092] providing a plurality of display elements;

[0093] generating a compressed video representation of each one of said elements; and storing said compressed representations in a memory, indexed by element.

[0094] Preferably, said display elements comprise GUI (graphical user interface) objects.

[0095] There is also provided in accordance with a preferred embodiment of the invention, a method-of-generating a catalog, comprising:

[0096] receiving, during interaction with an Internet, a plurality of objects to be displayed;

[0097] generating a compressed video representation of each one of said objects; and

[0098] storing said compressed representations in a cache of objects, whereby when one of said objects is received again from said Internet at a later time, a earlier stored compressed representation is used for display generation.

[0099] Preferably, said stored objects comprise at least one JPEG image. Alternatively or additionally, said stored objects comprise at least one uncompressed image. Alternatively or additionally, said stored objects comprise a WWW page.

[0100] There is also provided in accordance with a preferred embodiment of the invention, a method of Internet browsing, comprising:

[0101] connecting to an Internet;

[0102] receiving over said Internet at least one object to display;

[0103] retrieving a compressed representation of said object from a catalog; and

[0104] generating a compressed video stream utilizing said compressed representation.

[0105] Preferably, said object comprises a WWW page.

[0106] There is also provided in accordance with a preferred embodiment of the invention, a method of generating a compressed video stream, comprising:

[0107] generating a plurality of display commands which represents a display, by a program;

[0108] modifying at least one of said display commands, independently of said program;

[0109] generating a compressed video stream utilizing said display commands.

[0110] Preferably, said modifying comprises modifying said commands to responsive to limitations of a display device on which said compressed video stream is to be displayed. Alternatively or additionally, modifying comprises modifying said commands to reduce bandwidth requirements of said compressed video stream. Alternatively or additionally, modifying comprises modifying said commands to reduce resource requirements for compressing said compressed video stream. Alternatively or additionally, at least one of said display commands represents an object and said modifying comprises changing said at least one display command such that the object is moved relative to its original display position.

[0111] Preferably, said compressed video stream comprises utilizes blocks having boundaries and wherein moving comprises moving said object to match at least one compression block boundary.

[0112] In a preferred embodiment of the invention, at least one of said display commands represents an object and said modifying comprises replacing said object with a different object.

[0113] Preferably, replacing said object comprises replacing said object with a compressed representation of said different object. Alternatively or additionally, said object comprises a text object. Alternatively or additionally, said object comprises a background of said display. Alternatively or additionally, the method comprises analyzing said object to determine a closest suitable replacement object.

[0114] In a preferred embodiment of the invention, modifying said object comprises changing a font definition for said object. Alternatively or additionally, modifying said object comprises modifying at least one color of said object. Alternatively or additionally, modifying a color comprises reducing a spatial resolution of said colors. Preferably, modifying a color comprises reducing a color range resolution of said colors.

[0115] Alternatively or additionally, modifying said object comprises reducing a spatial resolution of said object.

[0116] In a preferred embodiment of the invention, at least one of said display commands comprises a scrolling command and wherein said modifying comprises increasing a granularity of said scrolling. Preferably, said increasing a granularity comprises limiting said scrolling command to multiples of compression blocks size of said compressed stream.

[0117] There is also provided in accordance with a preferred embodiment of the invention, a method of generating a compressed video stream, comprising:

[0118] generating a plurality of display commands which represents a display;

[0119] generating a plurality of transform coefficients from said display commands, wherein said transform coefficients are quantized and wherein said quantization is modified responsive to limitations a display device on which said compressed video stream is to be displayed; and

[0120] creating a compressed video stream utilizing said coefficients.

[0121] Preferably, said coefficients are quantized at a lower resolution in a vertical direction of said display.

[0122] There is also provided in accordance with a preferred embodiment of the invention, a method of generating a compressed video stream, comprising:

[0123] providing a display definition including a compressed object; and

[0124] converting said display definition into a compressed video stream, wherein converting includes transcribing at least a portion of said compressed object, without decompressing said compressed object.

[0125] Preferably, transcribing comprises reducing a quality level of said compressed object. Alternatively or additionally, said compressed stream and said compressed object are compressed utilizing a same transform and transcribing comprises copying transform coefficients from said compressed object to said compressed video stream. Alternatively or additionally, said compressed stream and said compressed object are compressed utilizing a motion estimation representation and transcribing comprises copying motion estimation vectors from said compressed object to said compressed video stream. Alternatively or additionally, converting said display definition comprises directly converting said display representation into said compressed video stream, without generating an intermediate image raster. Alternatively or additionally, said compressed object comprises a second compressed video stream, covering only a portion of said display. Alternatively or additionally, said compressed object comprises a compressed image, covering only a portion of said display. Alternatively or additionally, said compressed object comprises a compressed audio stream.

[0126] There is also provided in accordance with a preferred embodiment of the invention, software program for WWW page design, comprising:

[0127] a computer readable media having stored thereon:

[0128] a restriction module which receives restrictions defining limitations imposed by a compression method to be used for displaying WWW pages; and

[0129] a design module, which lays out display elements, responsive to said received limitations.

[0130] Preferably, said restrictions include a block size definition. Alternatively or additionally, said restrictions include a bandwidth restriction, and wherein laying out of display elements comprises selecting display elements to match said bandwidth limitations. Alternatively or additionally, said media has stored thereon an automated WWW page generator for a WWW server. Preferably, the software comprises a communication module for receiving said restrictions from a server associated with said compression.

[0131] There is also provided in accordance with a preferred embodiment of the invention, a compression-modified software for performing at least one function and for generating at least one display, comprising:

[0132] a computer readable media having stored thereon:

[0133] a functional module which performs said function; and

[0134] a compression-responsive module which receives an indication of restrictions related to a compression of said display and which controls said module to generate a display responsive to said indication, wherein said display is modified relative to a display generated without said restrictions.

[0135] Preferably, said indication comprises a message from a computer on which said software is executed. Alternatively or additionally, said indication comprises a configuration file. Alternatively or additionally, said modified display is modified to meet a bandwidth requirement. Preferably, said bandwidth requirement is an instantaneous band-width requirement.

[0136] In a preferred embodiment of the invention, said modified display is modified to reduce resources required for compression. Alternatively or additionally, said display is modified by moving at least one object, relative to its display location for a non-compressed display. Alternatively or additionally, said display is modified by utilizing a different object for a compressed display than for a non-compressed display. Alternatively or additionally, said function comprises a WWW browsing function.

[0137] There is also provided in accordance with a preferred embodiment of the invention, a method of display control, comprising:

[0138] transmitting a request, from a generator of a compressed video stream, to a source of display commands, which requests relates to said compression; and

[0139] modifying a generation of display commands, at said source, responsive to said request.

[0140] Preferably, said source of display commands comprises a WWV browser.

[0141] There is also provided in accordance with a preferred embodiment of the invention, a method of image processing, comprising:

[0142] providing an image to be compressed for transmission;

[0143] simultaneously compressing and processing said image;

[0144] transmitting said image to a remote location; and

[0145] decompressing said processed image at said remote location.

[0146] Preferably, said simultaneously compressing and processing comprises processing said image by modifying a quantization of at least one transform coefficient of said image. Alternatively or additionally, said modifying comprises increasing a quantization granularity for vertical coefficients.

[0147] There is also provided in accordance with a preferred embodiment of the invention, a method of generating a compressed video stream, comprising:

[0148] accumulating a plurality of display commands; identifying at least one display command of said plurality, whose effect is mooted by a later accumulated display command; and

[0149] converting said commands into said compressed video stream, ignoring the mooted display command.

[0150] Preferably, said mooted display command is ignored if its effect on said compressed video stream, in view of said later command, is below a threshold value. Alternatively or additionally, converting said commands comprises combining display commands. Alternatively or additionally, said converting comprises directly converting said commands into said compressed video stream without first generating an image raster.

[0151] There is also provided in accordance with a preferred embodiment of the invention, a method of video compression, comprising:

[0152] generating a display, which generation is refreshed at a generation refresh frame rate;

[0153] compressing, transmitting and decompressing said display; and

[0154] displaying said display at a second refresh rate other than said generation refresh rate and wherein said refresh is not synchronized to said display refresh rate. Preferably, said compressing is un-synchronized with said generation refresh frame rate. Alternatively or additionally, said transmitting is un-synchronized with said compressing.

[0155] There is also provided in accordance with a preferred embodiment of the invention, a method of generating a compressed video stream, comprising:

[0156] generating a plurality of display commands;

[0157] accumulating at least one display command;

[0158] generating a compressed video stream from said accumulated display commands, wherein accumulating comprises accumulating a number of display commands responsive to instantaneous available resources.

[0159] There is also provided in accordance with a preferred embodiment of the invention, a method of generating a compressed video stream, comprising:

[0160] generating a plurality of display commands;

[0161] accumulating at least one display command;

[0162] generating a compressed video stream from said accumulated display commands, wherein accumulating comprises accumulating a number of display commands responsive to a desired output frame rate.

[0163] In a preferred embodiment of the invention, generating a compressed video stream is un-synchronized relative to said generation of display commands. Preferably, said display commands are generated in sets, each set associated with a display frame and wherein said accumulating, in a single accumulation set, display commands originating from different sets, wherein said single accumulation set is utilized for generating a single frame of said compressed video stream.

[0164] Alternatively or additionally, the method comprises assigning a priority to a display command. Preferably, the method comprises reordering said accumulated display commands, responsive to said priority, wherein generating a compressed video stream comprises generating a first frame in said compressed video stream utilizing an originally later display command, prior to generating a frame in said stream utilizing an originally earlier display command.

[0165] There is also provided in accordance with a preferred embodiment of the invention, a method of generating a compressed video stream, comprising:

[0166] generating a plurality of display commands;

[0167] generating a compressed image portion responsive to said display commands; and

[0168] assembling a compressed video stream from said compressed image portion and from at least one previously compressed image portion, created prior to generation of said display commands.

[0169] Preferably, said previously compressed image portion comprises an image portion created for a previous image of said video stream.

[0170] There is also provided in accordance with a preferred embodiment of the invention, a method of asynchronous video stream compression, comprising:

[0171] providing a first data source and a second data source;

[0172] converting said first data source into a first portion of a compressed video stream;

[0173] converting said second data source into a second portion of a compressed video stream; and

[0174] mixing said first portion and said second portions to create a single compressed video stream, wherein said first portion and said second portion temporally overlap in said single compressed video stream.

[0175] Preferably, said first portion and said second portion utilize different frame rates. Alternatively or additionally, said first portion and said second portion occupy different spatial areas in a display defined by said single compressed video stream. Alternatively or additionally, said first and said second data sources comprise sets of display commands. Preferably, converting said first data source and converting said second data source comprises directly converting said data sources without generating an intermediate image raster.

[0176] In a preferred embodiment of the invention, at least one of said first data source and said second data source comprises an image source. Alternatively or additionally, said first data source comprises a display portion affected by a command entered by a user who views said single compressed video source. Preferably, the method comprises identifying portions of a data display which are affected by said user command.

[0177] There is also provided in accordance with a preferred embodiment of the invention, a method of generating a compressed video stream, comprising:

[0178] executing a program, which program generates a set of display commands, representative of a display;

[0179] identifying changes in said display which are responsive to at least one type of user command; and

[0180] converting said display commands into a compressed video stream,

[0181] wherein said changes are inserted into said compressed video stream at a higher frame rate than other changes in said display.

[0182] Preferably, said type of user command comprises a pointing device command. Alternatively or additionally, said changes comprises an indication of a selection of a GUI (graphical user interface) element. Alternatively or additionally, the method comprises analyzing said user command to determine display commands which effect said identified changes.

[0183] There is also provided in accordance with a preferred embodiment of the invention, a method of frame rate buffering comprising:

[0184] providing a source of compressed video which generates a compressed video stream having a variable frame rate;

[0185] providing a video display unit which receives said compressed video frames, decompresses said video frames and displays said video frames, wherein said unit is constrained to a fixed frame rate; and

[0186] padding said generated compressed video frames with frames which indicate that no change has occurred, to achieve said fixed frame rate.

[0187] Preferably, the method comprises increasing said padding and decreasing said variable rate, to compensate for bandwidth limitation in transmission between said source and said display unit. Alternatively or additionally, the method comprises increasing said padding and decreasing said variable rate, to compensate for an instantaneous resource limitation at said source.

[0188] There is also provided in accordance with a preferred embodiment of the invention, a method of personalizing advertising, comprising:

[0189] providing a compressed video stream; and

[0190] replacing at least one spatial portion of said compressed video stream with compressed data representing at least one advertisement, wherein said replacement is responsive to information associated with a viewer of said video stream.

[0191] Preferably, providing a compressed video stream comprises converting a computer software generated display into said compressed video stream. Preferably, said converting comprises directly converting a sequence of display commands generated by said software into said compressed video stream. Alternatively or additionally, said advertisement replaces a display object of said display. Alternatively or additionally, said replacing comprises replacing without decompressing said compressed video stream.

[0192] There is also provided in accordance with a preferred embodiment of the invention, a method of advertisement, comprising

[0193] providing a compressed video stream, representing a display;

[0194] determining an available portion of said display; and

[0195] inserting an advertisement into said available portion of said display, wherein inserting comprises manipulating said compressed video stream, without decompressing said compressed video stream.

[0196] Preferably, determining an available portion comprises receiving at least one indication of said available portion. Preferably, providing a compressed video stream comprises retrieving said stream from an archive location, wherein said archive location includes a storage of said indications.

[0197] Alternatively or additionally, providing a compressed video stream comprises converting a computer software generated display into said compressed video stream. Preferably, said software provides said indication. Alternatively or additionally, the method comprises analyzing said display to determine said available portion. Preferably, analyzing said display comprises analyzing display commands generated by said software.

[0198] In a preferred embodiment of the invention, said software comprises a WWW browser. Preferably, analyzing said display comprises identifying display areas into which additional display material is being downloaded.

[0199] There is also provided in accordance with a preferred embodiment of the invention, a charge accumulation method, comprising:

[0200] tracking a plurality of display transmissions from a server generating compressed video streams to a display unit which displays said streams, which transmissions utilize a variable video stream quality;

[0201] determining, for each transmission, a quality level of the video stream; and

[0202] generating a charge for using said transmissions, utilizing said determined quality levels.

[0203] Preferably, said display transmissions comprise TV program transmission. Alternatively or additionally, said display transmission comprises a transmission of a computer generated display. Preferably, said computer generated display comprises a WWW browser display.

[0204] In a preferred embodiment of the invention, generating a charge comprises generating a charge responsive to a predetermined quality level.

[0205] There is also provided in accordance with a preferred embodiment of the invention, a method of interactive TV, comprising:

[0206] displaying, at on an interactive TV, a WWW page including indications for TV channels;

[0207] detecting an interaction of a user with one of said indications; and

[0208] displaying a TV channel on said interactive TV responsive to said detection of interaction.

[0209] Preferably, said TV channel comprises a pay-on-demand movie.

[0210] There is also provided in accordance with a preferred embodiment of the invention, a method of interactive TV, comprising:

[0211] providing a compressed video stream representing a TV channel;

[0212] overlaying on said compressed video stream an interaction layer, including at least one control;

[0213] receiving from a viewer of said video stream an interaction with said control, wherein said overlaying comprises overlaying a compressed interaction layer on said compressed video, without decompressing said compressed video; and

[0214] modifying said compressed video stream responsive to said received interaction.

[0215] There is also provided in accordance with a preferred embodiment of the invention, a method of bandwidth allocation for a compressed video stream, comprising:

[0216] generating a plurality of display commands, by executing a computer program;

[0217] converting said display commands into a compressed video stream;

[0218] estimating a future content of said video stream; and

[0219] allocating bandwidth resources responsive to said estimate.

[0220] There is also provided in accordance with a preferred embodiment of the invention, a method of bandwidth allocation for a compressed video stream, comprising:

[0221] generating a plurality of display commands, by executing a computer program;

[0222] converting said display commands into a compressed video stream;

[0223] estimating a future content of said video stream; and

[0224] allocating CPU resources for compression responsive to said estimate.

[0225] In a preferred embodiment of the invention, said program comprises a WWW browser. Preferably, estimating comprises identifying a future download of complex display data. Alternatively or additionally, estimating comprises identifying a future download of a continuous data stream.

[0226] There is also provided in accordance with a preferred embodiment of the invention, a method of bandwidth allocation for transmitting video on a cable network, comprising:

[0227] providing a plurality of data sources;

[0228] differentially converting said data sources into compressed video streams, responsive to an instantaneous resource restriction; and

[0229] multiplexing said compressed video streams on a single transmission line.

[0230] Preferably, said differentially converting comprises converting each data source to a different frame rate compressed video stream. Alternatively or additionally, said differentially converting comprises, converting each data source to a different frame quality level. Alternatively or additionally, said resource restriction comprises a bandwidth restriction. Alternatively or additionally, said resource restriction comprises a computing resource restriction. Alternatively or additionally, said data sources comprise display commands. Alternatively or additionally, said differentially converting comprises differentially converting responsive to a content of said data sources. Preferably, the method comprises providing an indication of said content with said data sources. Alternatively or additionally, the method comprises providing an indication of said content by analyzing display commands which are comprised in said data sources. Alternatively or additionally, the method comprises providing an indication of said content by a software which generates at least one of said data sources.

[0231] There is also provided in accordance with a preferred embodiment of the invention, a method of generating a plurality of displays, comprising:

[0232] generating a first set of display commands, by a first program;

[0233] generating at least a second set of display commands, by at least a second program;

[0234] differentially affecting said first and said second programs, to generate said display commands; and

[0235] converting each of said first and said second sets of display commands into a compressed video stream,

[0236] wherein said differentially affecting comprises differentially affecting to meet an instantaneous resource limitation.

[0237] There is also provided in accordance with a preferred embodiment of the invention, a method of generating a plurality of displays, comprising:

[0238] generating a first set of display commands, by a first program;

[0239] generating at least a second set of display commands, by at least a second program;

[0240] differentially modifying said first and said sets of display commands; and

[0241] converting each of said first and said second sets of display commands into a compressed video stream,

[0242] wherein said differentially modifying comprises differentially modifying to meet an instantaneous resource limitation.

[0243] In a preferred embodiment of the invention, said resource limitation comprises a transmission bandwidth limitation. Alternatively or additionally, said resource limitation comprises a limitation on CPU available to perform said conversion.

[0244] There is also provided in accordance with a preferred embodiment of the invention, a method of bandwidth allocation, comprising:

[0245] providing a distribution network having a bandwidth;

[0246] transmitting on said network a plurality of channels, comprising Internet channels and TV channels; and

[0247] dynamically allocating bandwidth between Internet channels and TV channels.

[0248] There is also provided in accordance with a preferred embodiment of the invention, a method of transmitting a plurality of similar compressed video channels, comprising:

[0249] transmitting a base compressed image stream on a first channel;

[0250] transmitting modifications to said base image stream on at least one second channel;

[0251] receiving, at a display location, said first and said second channel;

[0252] modifying said first channel utilizing said second channel; and

[0253] displaying said modified first channel at said display location.

[0254] Preferably, one channel of said at least one second channel is targeted for said display location. Alternatively or additionally, transmitting comprises transmitting over a satellite network. Alternatively or additionally, different display locations utilize different ones of said at least one second channel to modify said first channel. Alternatively or additionally, said first channel carries a TV program and wherein said at least one second channel carries advertisements. Alternatively or additionally, said first channel carries a WWW site and wherein said at least one second channel carries personalizations of said site. Alternatively or additionally, said personalization comprises a scrolling of an object in said site. Alternatively or additionally, the method comprises assigning a third channel for use as a base image channel for said display unit, responsive to an interaction with said WWW site. Alternatively or additionally, the method comprises modifying an interactivity level of said site responsive to an availability of channels.

[0255] There is also provided in accordance with a preferred embodiment of the invention, a method of statistical bit multiplexing, comprising:

[0256] providing a plurality of compressed video streams to be multiplexed;

[0257] providing, for at least one of said plurality of streams, side information, indicative of a content of a frame of said stream; and

[0258] differentially dropping bits from said at least one of plurality of streams, responsive to said side information.

[0259] Preferably, said side information includes a minimal quality level for said frame.

[0260] There is also provided in accordance with a preferred embodiment of the invention, a method of generating a plurality of unrelated image streams, comprising:

[0261] defining a virtual display on a computer;

[0262] executing a plurality of programs on said computer, each program outputting to a different section of said display;

[0263] capturing, form each program, display commands for said display; and

[0264] converting each of said sets of display commands into an image stream having a content unrelated to image streams converted from other of said sets of display commands.

[0265] Preferably, said image streams comprise compressed image streams.

[0266] There is also provided in accordance with a preferred embodiment of the invention, a method of generating a plurality of unrelated audio streams, comprising:

[0267] providing a computer;

[0268] defining at least one virtual audio output on said computer;

[0269] executing a plurality of programs on said computer, each program outputting to said at least one audio output;

[0270] capturing, for each program, audio commands for said at least one audio output; and

[0271] converting each of said sets of audio commands into an audio stream having a content unrelated to audio streams converted from other of said sets of audio commands.

[0272] There is also provided in accordance with a preferred embodiment of the invention, method of producing different display representations at a plurality of remote locations, comprising:

[0273] transmitting a digitally encoded representation of a base display to a plurality of remote locations;

[0274] transmitting a plurality of representations of modifications of said base display;

[0275] receiving at a plurality of display locations said base display and at least one representation of a modification; and

[0276] reconstructing at said display locations, said display representations, from said base representation and said at least one modification representations.

[0277] Preferably, said representations comprise HTML files. Alternatively or additionally, said representations comprise sets of display commands. Alternatively or additionally, said representations comprise programs in a display generation language. Preferably, the display generation language comprises JAVA.

BRIEF DESCRIPTION OF THE DRAWINGS

[0278] The invention will be more clearly understood by reference to the following description of preferred embodiments thereof in conjunction with the figures, wherein identical structures, elements or parts which appear in more than one figure are labeled with the same numeral in all the figures in which they appear, in which:

[0279] FIG. 1 schematically illustrates a client/server configuration, in accordance with a preferred embodiment of the invention;

[0280] FIG. 2 schematically illustrates a server, in accordance with a preferred embodiment of the invention;

[0281] FIG. 3 is a flowchart for a method of generating a compressed video stream, in accordance with a preferred embodiment of the invention; and

[0282] FIG. 4 is a schematic diagram of a cable operator configuration, in accordance with a preferred embodiment of the invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0283] FIG. 1 schematically illustrates a client/server configuration 10, in accordance with a preferred embodiment of the invention. A server 16
communicates with a client 11, using a compressed video signal. Typically, a single server 16 is connected to and services a large number of clients 11. Client 11 comprises a video display, preferably a television set 12 (TV). The compressed video signal is preferably received by a set-top box 14 which decompresses the compressed signal and sends a video signal to TV 12. In a preferred embodiment of the invention, set-top box 14 is also used for viewing cable channels and/or pay-per-view programs.

[0284] In a preferred embodiment of the invention, client 11 includes input devices 20, preferably a mouse and/or a key-board. Alternatively, a remote-control with a small number of buttons is used for input. Input devices 20 may be connected to server 16 via a cable downlink connection, such as through set-top box 14. Such connections are well known in the art and may be used for transmitting small amounts of information back to a cable server. One such connection is made by the NCI corporation (network Computer Inc.). Alternatively or additionally, the connection may be via a telephone connection 22. In a preferred embodiment of the invention, user information, such as images and e-mail letters are faxed to server 16.

[0285] In a preferred embodiment of the invention, input devices 20 are wireless, such as infrared or ultrasonic, and set-top box 14 or telephone connection 22 include a suitable detector.

[0286] In some preferred embodiments of the invention, client 11 may comprise a computer 24 which can perform both the decompression and the display. Alternatively or additionally, the compressed video stream is used to carry data, preferably on a data channel of the compression protocol, to computer 24, from server 16. Such data communications may additionally or alternatively lead from computer 24 directly to server 16. Alternatively or additionally, input devices of computer 24, such as a mouse, may be used as an input device 20.

[0287] In one preferred embodiment of the invention, a direct computer/server connection is used to upload files to server 16 for transmission elsewhere and/or for downloading data files from the Internet.

[0288] In a preferred embodiment of the invention, server 16 is connected to an Internet 18 and client 11 is used to browse the Internet. Alternatively, server 16 may run other programs, such as a word processor, a spread-sheet or a computer game. In a preferred embodiment of the invention, the program executed by server 16 is a program which manages a subscriber's interaction with the cable company. In another preferred embodiment of the invention, the program is an e-mail program.

[0289] In one preferred embodiment of the invention, a program executed on server 16 may be used to browse data stored at server 16, for example DVD multimedia, in a preferred embodiment of the invention, server 16
maintains a copy of information which is often downloaded from the Internet, so that it can be simply and quickly browsed. Such information preferably includes, games, images, on-line journals and/or and data which is often downloaded and/or downloaded by more than a certain number of users.

[0290] FIG. 2 schematically illustrates a server 16, in accordance with a preferred embodiment of the invention. A program 30 is run at server 16
and generates a display to be viewed. This display is preferably drawn in a window on a virtual display 32. The window portion of display 32 is compressed, preferably to comply with the MPEG II standard. This compression may be performed by a standard MPEG II compressor. However, in a preferred embodiment of the invention, a special MPEG compression algorithm is used which take into account the type of display being generated.

[0291] It should be appreciated that although an MPEG II compression is assumed in many of the examples described herein, the invention is not limited to being used with MPEG II compression. Rather, many other compression mechanisms may be used, including advanced versions of MPEG, Quick time, wavelet compression, AVI and propriety compression protocols used by some multimedia data compressors.

[0292] The compressed display is then preferably physically encoded using a transport encoder 36 and mixed into an output video stream using a mixing box 38. User input is preferably filtered out by mixing box 38 and then directed to program 30, in a manner which emulates natural keyboard and/or mouse input to the program. Alternatively or additionally, especially when a special input device 20 is used, for example a remote, a special device driver in server 16 converts the user input into a format which can be assimilated by program 30.

[0293] In a preferred embodiment of the invention, server 16 runs a plurality of programs 30', the displays of which are all compressed, encoded and mixed into a cable wire signal. Preferably, the plurality of programs all run on a single CPU. Alternatively or additionally, server 16 may comprise a multi-CPU platform. Preferably, all the programs share a single virtual display. Alternatively, at least some of the programs may utilize a separate virtual display. Preferably, MPEG compression processes for a plurality of programs are performed by the same CPU. Alternatively, server 16 may include a plurality of DSP cards, which can be dynamically assigned to perform the compression. Preferably, all the compressed displays are transport encoded using a single transport encoder, alternatively a plurality of encoders may be used. Preferably, such a plurality of encoders are dynamically assigned for a display frame. Preferably, the encoders are genlocked and/or frame locked.

[0294] In a preferred embodiment of the invention, the various programs independently utilize system resources of server 16. In some cases, an operating system, such as windows NT or UNIX may support multiple users on a single system. Alternatively or additionally, server 16 may include software which captures operating system requests from the different programs 30 and then handles the requests in a manner which is transparent to the programs. One example of such a resource is cursor position. Another example is copy and paste commands in which a separate cut-buffer is preferably maintained for each program and/or user, so that copy and paste commands on different programs do not interact.

[0295] In a preferred embodiment of the invention, an overlay generator 40
is used to add an overlay to virtual display 32. Such an overlay may be used, for example, to add a cursor. Alternatively or additionally, the overlay is used to add user specific information which is not known by the program but which is tracked by server 16, for example connection time.

[0296] In a preferred embodiment of the invention, overlay generator 40
and/or other elements of server 16 may be used to add an overlay to a TV program channel, especially a compressed channel. Such an overlay may include WWW pages, program outputs, links and/or dynamic and/or static displays.

[0297] In a preferred embodiment of the invention, each program 30 has assigned to it also a virtual audio card, for outputting sounds generated by program 30. Alternatively or additionally, each program 30 has assigned to it a virtual video card. Preferably, the video output of the card are displayed in a window on the display. Preferably, the inputs and/or outputs utilized by program 30 are also managed as virtual devices, for example a virtual mouse and/or keyboard. In a preferred embodiment of the invention, additional virtual input and/or output devices may be associated with each program. One example of such a device is a modem, for a user downloading information to program 30. Another example is a voice input channel, preferably with speech recognition. Preferably, the virtual devices are assigned from a pool, on a demand basis. Alternatively or additionally, a single virtual device is used for a plurality of programs and the device driver is operative to individually service a plurality of programs so that there is no undesired interaction between the programs.

[0298] In a preferred embodiment of the invention, a portion of virtual display 32 may be viewed on server 16, for monitoring purposes. Alternatively or additionally, such a portion may be viewed from a remote location, preferably, by a remote system manager. Alternatively or additionally, a plurality of displays for individual programs may be displayed simultaneously, preferably with each one being reduced in size and with a plurality of program windows being displayed on a single monitor. In a preferred embodiment of the invention, server 16 is operative to record audio, displays and/or inputs for a particular program or for a plurality of programs. Such recording is preferably initiated by a system manager or by a user of the program. Preferably, an indication of such initiation is made directly to server 16 and bypassing program 30.

[0299] In a preferred embodiment of the invention, server 16 is connected to a printer and/or video or audio recording devices, such as a DVD. Preferably, a user can print text or graphics and/or record a multi-media segment using these devices and the hard copy is preferably mailed and/or otherwise delivered to the user.

[0300] In some preferred embodiments of the invention, mixing box 38 is used to connect to different types of networks, in addition to or instead of a cable network. In one preferred embodiment of the invention, mixing box 38 is used to connect to a telephone line. Preferably, a single data channel is transmitted over a plurality of lines. Alternatively, compression algorithms adapted for telephone lines may be used. Alternatively or additionally, an ISDN line may be used.

[0301] Preferably, set-top box 14 is modified so that it can also connect to the same type of network that mixing box 38 is connected to. Alternatively an additional adapter box is used to adapt from one network to the other.

[0302] In a preferred embodiment of the invention, a computer is directly connected to the network instead of through a set-top box. In one preferred embodiment of the invention, a LAN or a WAN is used to connect server 16 and the computer. Preferably, the compressed video data is transmitted as data packets over the network.

[0303] FIG. 3 is a flowchart for a method of generating a compressed video stream, in accordance with a preferred embodiment of the invention. This flowchart is preferably repeated (50) for each display element, where a display element is a graphical primitive or object which is drawn by program 30 or by the operating system. Alternatively, an entire display is built and then compressed. Alternatively the display is processed on a block-by block basis and each display element is a block or a portion thereof. Preferably, the block size is 8.times.8, 16.times.16 pixels or another multiple of the MPEG block size. It should be appreciated that different parts of the flowchart may be applied with a different granularity. For example, some of the flowchart may be performed on a frame-by-frame basis, while other parts may be performed on a block-by- block, super-block-by-super-block or element-by-element basis.

[0304] In a preferred embodiment of the invention, all the steps of the flowchart are performed. However, in other preferred embodiments of the invention, fewer than all the steps may be performed, for example even only a single step may be performed.

[0305] If the display element is an encoded object (52), for example compressed video or audio, the compressed data is preferably transcribed (54) into the MPEG stream, rather than decompressing and re-compressing it. When the object is displayed on the virtual display using a WWW browser, a special viewer is provided for this type of multimedia object so that the data is transcribed to the MPEG stream instead of being decompressed, "displayed" and recompressed. When the object is displayed using the operating system of server 16, a similar transcription is performed. In a preferred embodiment of the invention, the resolution of the transcribed stream may also be controlled, for example, by re-quantizing DCT coefficients in MPEG and JPEG data. Preferably, the resolution is controlled to match bandwidth restrictions and/or quality restrictions.

[0306] Different transcriptions are preferably used for different types of compression, since the type and amount of reusable data is dependent on the type of compression. For JPEG compressed images, the DCT coefficients may be reused. Once the image is transmitted once, it does not change, so only motion vectors are required to describe it in future frames. MPEG compressed streams may be simply copied. Some types of video compression do not use the DCT transform but do contain motion estimation parameters which are reusable by the MPEG compression process. The MPEG standard supports several types of audio, which types can be simply copied from the input to the MPEG stream. Alternatively or additionally, to compressed streams, other types of encoded data may be transcribed, for example, the encoded data may contain display commands, which may be directly converted to compressed video, without rendering the commands, as described herein.

[0307] In a preferred embodiment of the invention, the display is adjusted for the TV display (56) and/or to make the compression more efficient (58). These two types of adjustments may interact, however, for simplicity of description they are described separately. A TV display has several characteristics which not as good as and/or different from a computer display, including, resolution, frame rate, CRT scanning path, flicker problems, aspect ratio, viewing distance color resolution and interlacing. The lower resolution typically requires increasing a font size used or using a more-readable font than required for a computer display. Flicker problems are typically caused by thin horizontal lines, which are preferably replaced by thick horizontal lines. Alternatively or additionally, low pass filtering is applied to the image, at least in the vertical dimension. In a preferred embodiment of the invention, such a low pass filtering is applied by reducing the quantization of the vertical frequency coefficients, after the image is transformed, at least of the high frequencies. It should be appreciated that filtering in a DCT space is not multiplication. However, such filtering may be approximated by multiplication. Preferably however, a look-up table is used to perform filtering. Most preferably, the filtering is performed during the quantization, preferably embodied in the quantization look-up table.

[0308] The aspect ratio difference is preferably solved by pixel interpolation. The increased viewing distance for TV is preferably adjusted for by increasing the font size. The color resolution of a TV per-se may be the same as a computer display. However, both the MPEG II standard and most TV transmission standards utilize a lower spatial resolution for color. Preferably, display elements which depend on color resolution for their discernability are enlarged and/or the colors used changed. In a preferred embodiment of the invention, the display is modified so that it is suitable for vision-challenged individuals. Preferably, such modification is personalized to the visual abilities of a particular user, which abilities may be associated with the user or which abilities may be entered by the user.

[0309] There is preferably a two-way interaction between computer generated displays and MPEG compression. The MPEG compression affects how the display will look and the content of the display affects the efficiency and speed of the MPEG compression. First, some types of displays elements require many more bits to compress than other types. Second, some types of display elements are degraded to a greater amount than other elements, by a same amount of compression. Third, some types of display elements are easier/faster to compress than others. In a preferred embodiment of the invention, display elements are modified so that they look better, are easier to use and/or faster to compress than the original display elements. Some examples include, manipulating colors so that the color components can be compressed to a greater degree; moving elements, such as bullets, so that they do not straddle block boundaries; using fonts with fewer high-frequency components; reducing flashing rates of colors and cursors and/or synchronizing them with each other and/or other display changes; reducing animation rates of animated objects; low-pass filtering, to reduce display resolution; and/or reducing scrolling resolution, for example to be limited to blocks and/or super-blocks, so that motion estimation only involves moving of whole blocks. In a preferred embodiment of the invention, a scrolling below a certain resolution is not shown and/or is rounded up to a nearest number of pixels divisible by 8.

[0310] In a preferred embodiment of the invention, color manipulation and/or other types of image adjustment, except possibly for aspect ratio correction, are not performed on multimedia segments which are downloaded from the Internet.

[0311] Preferably, some or all of these adjustments are performed by changing settings of program 30 and/or of the operating system. However, in some cases, these adjustments may be simpler to perform as a post-processing step during the display compression (58).

[0312] In a preferred embodiment of the invention, motion estimation is performed utilizing a knowledge of the structure and/or build-up of the display, instead of or in addition to analyzing consecutive frames of the display (60). Preferably, this knowledge is gleaned from the graphics commands generated by program 30. Alternatively or additionally, especially in an Internet browsing embodiment, this knowledge is gleaned from HTML, Java or java script commands downloaded from the Internet. Similarly, changes in a display which are described below as being performed by modifying display commands, may also be performed by modifying HTML files, possibly in a pre-processor to the browser.

[0313] In a preferred embodiment of the invention, special note is taken of scrolling commands, which define a motion vector for portions of the display. Text based displays usually involve a significant amount of scrolling. Even graphics based displays may contain a significant amount of scrolling, since the display is often only a window into a large underlying data space. Preferably, very large motion vectors are allowed, for example as long as 200 pixel long. Typically, the vectors will comprise integer numbers of pixels in the X and Y axis components. It should be appreciated that there are several types of "scrolling" type commands, including:

[0314] (a) scrolling of a portion of the display when a user enters text in a word processor;

[0315] (b) scrolling of a scrollable display element;

[0316] (c) moving a window and/or a display element on the display;

[0317] (d) horizontal and/or vertical motions of blocks of image information; and/or

[0318] (e) scrolling of an entire horizontal or vertical swath of a screen or of the entire screen.

[0319] In a particular example, when a WWW page is being downloaded, portions of the display are moved around as new images and their sizes are downloaded. In a preferred embodiment of the invention, portions of the display are moved using scrolling-type commands, so that only motion vectors need to be transmitted by the compression mechanism. In a preferred embodiment of the invention, some of the data may be scrolled to or from outside of the viewing area. Preferably, scrolled out data may be retained by the MPEG decompressor, for use when such data "moves" back into the displayed area.

[0320] In addition, determining motion vectors by image analysis can generally be performed more efficiently in most computer generated displays than in acquired image streams. When a section of a computer generated display moves, it is usually copied, so a perfect correspondence between the source and the destination can be expected and searched for. In a preferred embodiment of the invention, the computer program is limited to movements of units of 8 or some other number, preferably integer, so that motion estimation is faster and/or does not required pixel interpolation. Alternatively or additionally, the virtual device driver translates scrolling commands into units which are divisible by 8.

[0321] In a preferred embodiment of the invention, changes are detected by focusing on pixels which were rendered and or on areas on which display commands operated (62).

[0322] In a preferred embodiment of the invention, graphical elements are drawn directly (64) into a transform space, such as a DCT space or a wavelet transform space, instead of first being drawn and then transformed. Some graphical elements can be reused (their DCT coefficients reused), for example, window borders, standard buttons, bullets, letters, icons, menus and/or some images. Graphical elements can be reused between downloaded web pages, between programs, between users and/or between different frames of the same program. Preferably, a cache of DCT transformed web pages is stored in addition to or instead of a cache of HTML files. In addition, basic graphical elements can be drawn directly as DCT coefficients, for example, pixels, line, combinations of primitives and transformed primitives (zoom, rotate).

[0323] In some cases there is a simple analytic formula connecting the original DCT coefficients and the transformed ones, for example in rotations, zoom by an integer number and decimation by an integer factor. When two primitives are both drawn in the same block, the coefficients are added. In a preferred embodiment of the invention, the coefficients are linearly added. Alternatively or additionally, the adding takes into account the quantization and is preferably performed using a look-up table. In a preferred embodiment of the invention, when the primitives overlap, they are split into non-overlapping portions and overlapping portions. The non-overlapping portions are simply drawn. The drawing of the overlapping portions depends on the type of interaction (such as bitblt command) between the two primitives. In some cases, a sum or an average of their DCT coefficients generates a good graphical representation.

[0324] Alternatively or additionally, other processing may be performed on DCT coefficients, prior to them being drawn into said DCT space, for example, rotation, quantization, smoothing and/or thresholding.

[0325] In a preferred embodiment of the invention, the DCT coefficients for some primitives are calculated on-the-fly, using analytical formulae. In one example, a single pixel, DCT coefficients for a pixel at a location (x0, y0) in a compression block yields the following DCT coefficient value for coefficients (k, i): cos(2.pi.0*(k+0.5)/8)cos(2.pi.- y0(i+0.5)/8). A line is preferably drawn as a series of pixels, each of which may be "rendered" using the above formula. Preferably, a line is first divided into line segments, each of which is included in a single compression block.

[0326] Alternatively or additionally, the DCT coefficients for basic objects are stored in tables, corresponding to positions of the object within the block. Preferably, display elements are modified so that they match a table entry and/or an analytic transformation is used to modify them. It should be appreciated that in some preferred embodiments of the invention, standard display portions, such as borders and menus change rarely on the display, so that they are only infrequently compressed and/or retrieved from an appropriate lookup table.

[0327] In a preferred embodiment of the invention, display elements are moved so that they do not straddle compression blocks. Alternatively, or when the element is larger than a block, the element may be split between blocks. Alternatively or additionally, a look-up table stores coefficients for a 16.times.16 block. Alternatively or additionally, the 16.times.16 coefficients are calculated from 8.times.8 coefficients, by interpolation.

[0328] It should be appreciated that methods for direct drawing of elements into the DCT space and/or other transform spaces, for other compression methods, are generally suitable for parallel implementation, as the different coefficients can be set independent of each other.

[0329] Up to this stage, the transformation, as described, is generally loss-less, although lossy transformation (compression) methods and/or thresholds can be applied even at these early stages. Information compression is preferably achieved by quantization of the DCT coefficients. It should be appreciated that in many implementation, the transform step is integrated with the quantization step, so there may be no separating line between transformation and compression. In a preferred embodiment of the invention, a compression depth which requires lossy compression is determined (66). Preferably, the depth of compression is determined responsive to external requirements (67), such as available bandwidth or others, described below. Alternatively or additionally, the amount of lossy compression desired for the current display is outputted to an external environment, preferably as part of a negotiation for bandwidth. The desired compression depth may be dependent on the type of data displayed or on other parameters, described below.

[0330] Additionally or alternatively to determining the compression depth, the frame rate of the display is reduced to below a video rate (68).

[0331] In a preferred embodiment of the invention, a plurality of modifiable parameters are available for trading off image quality and bandwidth. Each of these parameters may have a range in which the image quality is minimally degraded and a range where the quality is significantly degraded. In addition, different parameters may have different effects for different images types. For example, the frame rate is not very important when viewing text, but it is very important when playing a computer game. The parameters which can be modified preferably include one or more of:

[0332] (a) frame rate;

[0333] (b) DCT coefficient sampling resolution;

[0334] (c) a motion estimation threshold, below which no motion is detected;

[0335] (d) a change estimation threshold;

[0336] (e) allowances for modifying the positions of objects, for more efficient compression;

[0337] (f) allowance for modifying the appearance of objects;

[0338] (g) color modifications;

[0339] (h) reduced resolution in imported multimedia objects, such as MPEG movies;

[0340] (i) accuracy of direct DCT drawing tables and transformations;

[0341] (j) different parameter values per portion of the display, per display element and/or graphical primitive; and/or

[0342] (k) a desired image quality.

[0343] In a preferred embodiment of the invention, different parameters have different levels of importance, depending, inter alia, on what is being viewed. Preferably, each program, TV set, user and/or combinations thereof are associated with a particular set of desired values for these parameters and their relative importance and/or minimal and/or maximal values. Preferably, these parameters are associated with a set-top box digital subscriber number.

[0344] Additionally or alternatively, different values of compression parameters may be defined for different parts of the display, for example, standard icons and menu bars may be compressed to a lower quality than unknown icons, since they are more easily recognized.

[0345] Additionally or alternatively, the compression and/or bandwidth requirements are dependent on a prediction of the future frames. In one example, if a JPEG image is being downloaded and displayed, less bandwidth will be required for a particular display portion than if an MPEG video is being downloaded and displayed.

[0346] In a preferred embodiment of the invention, the compression is performed using a variable-bit rate, while requiring a constant quality threshold. The quality threshold may be changed as a function of available bandwidth, as described above.

[0347] In a preferred embodiment of the invention, the compressed video signal may include two types of compressed data. For example, an MPEG portion and an AVI portion. However, this may require additional capabilities for the set-top box.

[0348] In a preferred embodiment of the invention, programming of the set-top box is updated from server 16. This updating is preferably performed while the TV is not in use. Preferably, the selection of the compression parameters is a function of the TV type (standard, digital or HDTV), its size and/or the capabilities of the set-top box with regard to its speed, programmability and memory. Preferably, more complex and/or efficient compression schemes are used for more sophisticated set-top boxes.

[0349] Although the above embodiments have been described generally with reference to MPEG compression, it should be appreciated that they may be applied for many types of compression practiced today, especially transform-quantize type compression methods. Non MPEG compression types are also important when the display is not a two-dimensional raster display, for example, when the display is a three-dimensional raster display.

[0350] In a preferred embodiment of the invention, the method explained with reference to FIG. 3 is performed responsive to change events, which indicate to the image compressor that a change has occurred in an image of a display channel. Step 62, "change detection" is preferably utilized to determine if the change warrants an update to the image, based on, for example, the available bandwidth, available computing power and/or type of user connection. In a preferred embodiment of the invention, only the changed portion of the image is processed and a new output frame is assembled from the newly compressed image portions and, possibly, existing, unchanged compressed image portions. In a preferred embodiment of the invention, three kinds of frames may be generated:

[0351] (a) "I" frames, which in