Model-based view extrapolation for interactive virtual reality systems

• US 6,307,567 B1
• Filed: 12/22/1997
• Issued: 10/23/2001
• Est. Priority Date: 12/29/1996
• Status: Expired due to Term

• Current Assignee: Hanger Solutions, LLC (IP Investments Group LLC)
• Original Assignee: Richfx Incorporated
• Inventors: Cohen-or, Daniel
• Primary Examiner(s): Vo, Cliff N.
• Application Number:
  US08/995,568
• Time in Patent Office: 1,401 Days
• Field of Search: 345/433, 345/419, 345/474, 345/355, 345/357, 345/329
• US Class Current: 345/619
• CPC Class Codes: G06T 15/10 Geometric effects
• CPC Class Codes: H04N 19/20 using video object coding
• CPC Class Codes: H04N 19/503 involving temporal prediction adaptive coding with adaptive selection between spatial and temporal predictive coding H04N19/107; adaptive coding with adaptive selection among a plurality of temporal predictive coding modes H04N19/109

Abstract

A method by which a client and a server cooperate to render views of a virtual reality world within which a user of the client navigates. The server sends the client a reference view and a model to use to extrapolate the reference view in response to the user'"'"'s virtual movement. The client sends the server a record of the user'"'"'s virtual movement. Both the client and the server perform the model-based extrapolation of the reference view. The server also performs an exact update of the reference view. When the extrapolated reference view deviates too much from the updated reference view, the server sends the client the difference between the extrapolated and updated reference views, so that the client can update its reference view.

Claims

1. In a system in which a server and a client cooperate to render a plurality of views of a virtual world, each view corresponding to a viewpoint, a method for generating the views in real time, the method comprising:

(a) transmitting a first reference view to the client;

(b) transmitting at least a portion of a virtual world model to the client;

(c) extrapolating said first reference view based on said portion of said virtual world model to generate an extrapolated view;

(d) transmitting at least one correction dataset to the client; and

(e) correcting said extrapolated view based on said at least one correction dataset to generate at least one second reference view.

2. The method of claim 1, wherein said extrapolating comprises:

(i) transforming said virtual world model to the viewpoint of said extrapolated view;

(ii) rendering said virtual world model; and

(iii) backprojecting said virtual world model to said first reference view.

3. The method of claim 1, further comprising:

(f) rendering an exact view, by the server; and

(g) subtracting said extrapolated view from said exact view to generate a difference image for use as said at least one correction dataset.

4. The method of claim 3, further comprising:

(h) compressing said difference image.

5. The method of claim 1 further comprising:

(f) replacing said first reference view with said at least one second reference view.

6. The method of claim 1, wherein said correction dataset includes a difference image.

12. The method of claim 1, wherein said extrapolating of said first reference view is performed a plurality of times to generate a plurality of extrapolated views, and wherein said correcting is performed on only a portion of said plurality of extrapolated views.

7. In a system in which a server and a client cooperate to render a plurality of views of a virtual world, a method for updating the views in real time, the method comprising:

(a) transmitting a first reference view to the client;

(b) extrapolating said first reference view, to generate an extrapolated view;

(c) transmitting at least one correction dataset to the client; and

(d) correcting said extrapolated view based on said at least one correction dataset, to generate at least one second reference view;

wherein said extrapolating occurs at least twice before said correcting.

8. The method of claim 7, further comprising:

(e) rendering an exact view, by the server; and

(f) subtracting said extrapolated view from said exact view to generate a difference image for use as said at least one correction dataset.

9. The method of claim 8, further comprising:

(g) compressing said difference image.

10. The method of claim 7, further comprising:

(e) replacing said first reference view with said at least one second reference view.

11. The method of claim 7, wherein said correction dataset includes a difference image.

13. In a system for rendering a plurality of views, each of the views associated with a viewpoint, a method for reconstructing at least one of the plurality of views, the method comprising:

(a) providing one or more reference views;

(b) providing range data;

(c) extrapolating said one or more reference views based on said range data and on at least one of the viewpoints to generate at least one extrapolated view;

(d) providing at least one correction dataset; and

(e) correcting said at least one extrapolated view based on said at least one correction dataset to generate at least one new reference view.

14. The method of claim 13, wherein for at least one of said reference views, said extrapolating is performed a plurality of times to generate plurality of extrapolated views, wherein only a portion of said plurality of extrapolated views are corrected.

15. The method of claim 13, wherein said range data is provided as part of a geometric model.

16. The method of claim 13 further comprising:

(f) adding said new reference view to said one or more reference views.

17. The method of claim 13, wherein said extrapolating includes motion compensation.

18. The method of claim 17, wherein said motion compensation includes backprojection.

19. The method of claim 13, wherein said correction dataset comprises a difference between an exact view and said extrapolated view.

20. The method of claim 19, further comprising:

(f) rendering said exact view, based on a virtual reality world.

21. The method of claim 13, further comprising

(f) compressing said correction dataset.

22. The method of claim 21, wherein said compressing comprises using a lossy compression.

23. The method of claim 22, wherein said lossy compression is JPEG.

24. The method of claim 13, wherein only enough of said range data is provided to enable said extrapolating.

25. The method of claim 13, wherein said plurality of views are rendered cooperatively by a sender and a receiver.

26. The method of claim 25, wherein said providing of said one or more reference views, said providing of said range data, and said providing of said at least one correction dataset comprise sending said one or more reference views, said range data and said at least one correction dataset from said sender to said receiver.

27. The method of claim 25, wherein said extrapolating and said correcting are performed both by said sender and by said receiver.

28. The method of claim 25, wherein said receiver includes a client that includes a virtual camera, said at least one viewpoint being provided by said virtual camera, said range data being provided in accordance with said at least one viewpoint.

29. The method of claim 28, wherein said range data is included in a portion of a geometric model sent from said sender to said receiver, said portion of said geometric model being selected according to said at least one viewpoint.

30. The method of claim 28, wherein said range data is included in a portion of a geometric model sent from said sender to said receiver, said portion of said geometric model being selected according to a required level of detail.

31. The method of claim 25, wherein said sender includes a server and wherein said receiver includes a client, said server and said client being connected via a network.

32. The method of claim 31, further comprising:

(f) establishing at least one of the viewpoints, by said client.

33. The method of claim 32, further comprising:

(g) providing said at least one viewpoint to said server, by said client; and

(h) determining said at least one correction dataset based on said at least one viewpoint, by said server.

34. The method of claim 32, wherein said establishing of said at least one viewpoint is performed in accordance with user navigation.

35. In a network comprising a client and a server, a method performed at the client of rendering views of a virtual world, the method comprising:

generating an extrapolated view of the virtual world based on a reference view of the virtual world and on at least a portion of a virtual world model;

receiving from the server a correction dataset defining a difference between the extrapolated view of the virtual world and an exact view of the virtual world maintained at the server; and

modifying the extrapolated view based on the received correction dataset to generate an updated reference view to be used for generating subsequent extrapolation views of the virtual world.

36. The method of claim 35 in which generation of the extrapolated view of the virtual world is based at least in part on range data received from the server.

37. The method of claim 35 further comprising receiving from the server the virtual world reference view or the virtual world model portion or both.

38. In a network comprising a client and a server, a method performed at the server to facilitate rendering views of a virtual world at the client, the method comprising:

maintaining an exact view of the virtual world based on a virtual world model and a texture map residing at the server;

maintaining an extrapolated view of the virtual world based on a reference view and a subset of the virtual world model known to the client; and

transmitting to the client a correction dataset defining a difference between the extrapolated view of the virtual world and an exact view of the virtual world.

39. The method of claim 38 in which maintenance of the exact and extrapolated view s of the virtual world is based at least in part on user-input received from the client.

40. The method of claim 39 in which the user-input received from the client corresponds to a user'"'"'s viewpoint in the virtual world.

41. The method of claim 38 further comprising updating the reference view based on the correction dataset transmitted to the client.

42. The method of claim 41 further comprising generating subsequent extrapolated views based on the update reference view.

43. The method of claim 38 in which maintenance of the extrapolated view at the server comprises emulating a corresponding maintenance of an extrapolated view at the client.

44. Computer software, residing on a computer-readable medium, for rendering views of a virtual world in a network comprising a client and a server, the software comprising instructions for causing the client to perform the following operations:

generate an extrapolated view of the virtual world based on a reference view of the virtual world and on at least a portion of a virtual world model;

receive from the server a correction dataset defining a difference between the extrapolated view of the virtual world and an exact view of the virtual world maintained at the server; and

modify the extrapolated view based on the received correction dataset to generate an updated reference view to be used for generating subsequent extrapolation views of the virtual world.

45. Computer software, residing on a computer-readable medium, for rendering views of a virtual world in a network comprising a client and a server, the software comprising instructions for causing the server to perform the following operations:

maintain an exact view of the virtual world based on a virtual world model and a texture map residing at the server;

maintain an extrapolated view of the virtual world based on a reference view and a subset of the virtual world model known to the client; and

transmit to the client a correction dataset defining a difference between the extrapolated view of the virtual world and an exact view of the virtual world.