Method of and apparatus for constructing an image of a notional scene by a process of ray tracing

• US 5,933,146 A
• Filed: 05/28/1997
• Issued: 08/03/1999
• Est. Priority Date: 12/01/1994
• Status: Expired due to Fees

• Current Assignee: ART VPS Ltd.
• Original Assignee: Advanced Rendering Technology Ltd. (ART VPS Ltd.)
• Inventors: Wrigley, Adrian
• Primary Examiner(s): Powell, Mark R.
• Assistant Examiner(s): MANCUSO, HUEDUNG XUAN CAO
• Application Number:
  US08/849,596
• Time in Patent Office: 797 Days
• Field of Search: 345/420, 345/421, 345/422, 345/426
• US Class Current: 345/420
• CPC Class Codes: G06T 15/06 Ray-tracing
• CPC Class Codes: G06T 15/40 Hidden part removal
• CPC Class Codes: G06T 15/50 Lighting effects

Abstract

A ray tracing method for construction an image of a notional scene from data concerning a plurality of objects in the scene and the illumination of the scene, involves allocating surfaces of objects in the scene into a plurality of subsets, projecting from a notional viewing position a plurality of straight line paths and storing data for all said paths, determining which, if any, of those straight line paths intersect any surface in a selected subset or a zone of illumination, and repeating that step for all the other subsets until all the paths have been tested for intersections against all surfaces in the scene. The method also determines, for each path, which intersection is nearest the viewing position and stores data on the co-ordinates for that intersection and on the surface of the object or zone of illumination at that intersection. The intensity in colour of light which will pass along the path through the notional viewing position from that intersection is then calculated and stored, and the stored intensity and/or colour is then used to produce a picture signal for a visual display. A plurality of paths are tested for intersections against a given subset of surfaces before the paths are considered in relation to other subsets, thereby reducing the number of times the subsets have to be called up from the database. The method can be conducted on apparatus which uses parallel ray processors.

Claims

1. A method of constructing an image of a notional scene, containing a plurality of objects, as viewed from one of a plurality of possible viewing positions and illuminated in a known manner, from information on the co-ordinates of the set of surfaces defining the objects, on the optical characteristics of all the points on those surfaces and on the particular illumination, which information is stored in a computer memory, the method comprising the steps of:

A. Allocating the surfaces into a plurality of subsets each containing one or more surfaces;

B. Projecting from the viewing position a plurality of straight line paths and storing data describing all said paths;

C. For one of the subsets, determining which, if any, of the straight line paths intersect at least one of the surfaces in that subset, or a zone of illumination and determining the co-ordinates of said intersections;

D. Repeating step C above for the other subset or for each other subset in turn;

E. Determining, for each path which intersects a surface or zone of illumination, the nearest intersection to the viewing position and storing same with data on the object surface or zone of illumination at that intersection;

F. Determining, for each nearest intersection the light which will pass along the path to the viewing position from said intersection and storing same as light intensity and position data; and

G. Reading out the stored light intensity and position data, to produce a picture signal for a visual display.

2. A method according to claim 1 in which said plurality of straight line paths are of a sufficient number to enable the whole of the image to be constructed.

3. A method according to claim 1 in which step F involves, for each nearest intersection with a surface, generating a set of one or more secondary straight line paths, from an origin defined by said nearest intersection, if the characteristics of the surface at the origin are such that light incident thereon along the or each secondary path will be directed to said viewing position, determining the intersections of the or each secondary path with other surfaces or zones of illumination closest to said origin in accordance with steps C and D and calculating at least the intensity of light incident on the origin along each secondary path, and calculating, from the characteristics of the surface at the origin, at least the intensity of that light which is incident on the viewing position from the origin.

4. A method according to claim 3 in which step F involves the further steps of generating one or more further sets of straight line paths, in a similar fashion to the secondary straight line paths, from the origins defined by selected intersections of the previous set of straight line paths and determining the intensity of light incident on those origins along directions defined by said further set of paths.

5. A method according to claim 1 in which the determining of which straight line paths intersect a selected surface includes the step of determining whether each path intersects a notional envelope which contains said surface, and discarding the path from further consideration in relation to the surface if it does not intersect the envelope.

6. A method according to claim 5 in which the number of paths which need to be considered in relation to the envelope is reduced by determining which paths intersect a larger envelope which contains a said envelope, and discarding paths which do not intersect said larger envelope from consideration in relation to the envelope.

7. A method according to claim 3 in which, where the surface characteristic of an object at said origin is such that incident light thereon would be diffused or scattered in a multitude of directions, so that light from each of a multitude of further paths would be directed along said previous path, only one such further path is generated at random.

20. A method according to claim 3 in which each set of secondary or further set of straight line paths is generated when an intersection of the previous straight line path is found, and before the intersection of the previous path nearest its origin or, as the case may be, the viewing position is determined, and each set of secondary or further paths which does not originate from said nearest intersection is subsequently discarded.

8. Apparatus for constructing an image of a notional scene containing a plurality of objects as viewed from one of a plurality of possible viewing positions, the apparatus comprising:

memory means on which information on the co-ordinates of the surfaces of the objects, the optical characteristics of all the points on those surfaces and on the illumination for the scene is stored, the co-ordinate data being arranged in subsets each containing data for one or more object surfaces;

ray tracing means for projecting from the viewing position a plurality of straight line paths for determining which, if any, of the paths intersect a surface of each object or a zone of illumination, and for selecting the intersections of each path closest to the viewing position;

calculating means for determining, from the data on the object surfaces at said intersections and on the illumination of the scene, at least the intensity of the light which will pass along each said path to the viewing position; and

output means for generating a picture signal, representative of said intensities, for display on a visual display, wherein all the intersections for a first subset of one or more object surfaces selected by the ray tracing means are determined before the intersections with the or each other subset, so that the intersections are determined in respect of each subset in turn.

9. Apparatus according to claim 8 in which the ray tracing means comprises a plurality of data processors which are all connected to the memory, the arrangement being such that each processor generates a respective group of straight line paths for a respective part of the image, the memory broadcasts each subset in turn to the processors, and the intersections of each group with the or each surface of a selected subset are determined concurrently.

10. Apparatus according to claim 9 in which each processor comprises a dedicated chip specifically configured to perform the ray tracing operations.

11. Apparatus according to claim 10 in which each processor is so arranged as to perform different operations on different paths, in its respective group, concurrently.

12. Apparatus according to claim 8 in which the calculating means is so arranged so as to combine the effects of the interactions of a notional light ray at the relevant interfaces by a process of logarithmic arithmetic.

13. Apparatus according to claim 8 in which the ray tracing means is operable to create a logarithmic representation of the co-ordinates of each straight line path and to test each path for intersections and determine the co-ordinates of the nearest intersections by means of logarithmic calculations.

14. Apparatus according claim 9 in which each ray tracing processor is arranged to test its respective group of straight line paths for intersections with notional envelopes which contain the surface or surfaces in respect of which said intersection tests are to be conducted, and the apparatus has storage means for storing data identifying the paths which have intersected the envelope so that further intersection tests can subsequently be performed by the processor on those paths.

15. Apparatus according to claim 14 in which the storage means is also arranged to store the co-ordinates of said envelopes, and comprises part of said memory means.

16. Apparatus according to claim 9 in which said processors also incorporate the calculating means so that each processor is also operable to calculate the intensity of light travelling along each of its paths to the viewing position.

17. Apparatus according to claim 16 in which each processor includes or is connected to a respective texture memory unit which is connected to the memory means, and which is operable temporarily to store a texture map of surface characteristics of the surface or surfaces in the scene within a region defined by the boundary of the map for access by the calculating means to enable the latter to calculate the light intensities and/or colours in respect of the straight line paths intersecting the surfaces in said region.

18. Apparatus according to claim 17 in which the memory means is operable to overwrite any texture map currently stored in each texture store to enable the calculating means to calculate the light intensities and/or colours in respect of straight line paths which intersect surfaces in a different region of the scene.

19. Apparatus according to claim 17 in which the processor is operable to determine which paths intersect a texture map on a surface by dividing the surface into one or more regions and determining which paths intersect a region which contains the co-ordinates of at least part of the map and then determining which of these paths intersect the region defined by the co-ordinates of the map.

21. A method of constructing an image of a notional scene, containing a plurality of objects, as viewed from one of a plurality of possible viewing positions and illuminated in a known manner, from information on the co-ordinates of the set of surfaces defining the objects, on the optical characteristics of all the points on those surfaces and on the particular illumination, which information is stored in a computer memory, the method comprising the steps of

A. allocating the surfaces into a plurality of subsets each containing a respective group of surfaces;

B. projecting from the viewing position a plurality of straight line paths and storing data describing all said paths;

C. for one of the subsets, determining which, if any, of the straight line paths intersect at least one of the surfaces in that subset, or a zone of illumination and determining the co-ordinates of said intersections;

D. repeating step C above for the other subset or for each other subset in turn;

E. determining, for each path which intersects a surface or zone of illumination, the nearest intersection to the viewing position and storing same with data on the object surface or zone of illumination at that intersection;

F. determining, for each nearest intersection the light which will pass along the path to the viewing position from said intersection and storing same as light intensity and position data; and

G. reading out the stored light intensity and position data, to produce a picture signal for a visual display,

wherein all the paths are generated before intersections with the surfaces are determined, all the paths are tested for intersections with all the surfaces in a given subset before any one of the paths is tested for intersections with the surfaces in the next subset and wherein information on the position and shape of each surface is stored in memory means from which information on the surface is retrieved only once in order to determine intersections of that surface with all the paths.