Image generator for generating perspective views from data defining a model having opaque and translucent features

• US 5,363,475 A
• Filed: 11/08/1994
• Issued: 11/08/1994
• Est. Priority Date: 12/05/1988
• Status: Expired due to Fees

• Current Assignee: Rediffusion Simulation Ltd (Thales SA)
• Original Assignee: Rediffusion Simulation Ltd (Thales SA)
• Inventors: Baker, Stephen J., Wood, Karl J., Olive, Graham J., Cowdrey, Dennis A.
• Primary Examiner(s): Zimmerman, Mark K.
• Application Number:
  US07/689,924
• Time in Patent Office: Days
• Field of Search: 395/122, 395/126, 395/131, 395/164
• US Class Current: 345/422
• CPC Class Codes: G06T 1/60 Memory management
• CPC Class Codes: G06T 15/00 3D [Three Dimensional] image rendering
• CPC Class Codes: G09B 9/301 by computer-processed or -generated image

Abstract

An apparatus for generating an image from data defining a model including a plurality of opaque and translucent features. The image is intended to represent a view of the model from a predetermined eyepoint and is made up from an array of screen space pixels. The image area is divided into an array of sub-areas each of which covers at least one pixel. For each feature in the model that is potentially visible from the eyepoint, a test is conducted to determine which of the sub-areas is at least partially covered by that feature. For each feature which covers a sampling point, a function of the distance from the eyepoint to that feature at the sampling point is determined. An output for each pixel within a sub-area is produced, the pixel output corresponding to the combined effects of the sampling point outputs for all sampling points which contribute to that pixel, and the pixel outputs are displayed.

Claims

1. An apparatus for generating an image to be displayed on a display screen from data defining a model including a plurality of opaque and translucent features, the image being intended to represent a view of the model from a predetermined eyepoint and being made up from an angry of screen space pixels to be displayed by a raster scanning process, each pixel being of uniform color and intensity, and the pixels together defining an image area, comprising:

a. dividing means for dividing the image area into an array of sub-areas each of which covers at least one pixel,

b. sub-area coverage determining means for determining for each feature in the model which of the sub-areas is at least partially covered by that feature,

c. list means for producing a list of feature identifiers in respect of each sub-area, the list for any one sub-area identifying features which at least partially cover that sub-area,

d. position means for determining a position in screen space for at least one sampling point within each sub-area,

e. sampling point cover determining means for determining, for each sub-area in turn, and for each said sampling point, which of the features in that sub-area'"'"'s list cover that sampling point,

f. distance determining means for determining, for each feature which covers a sampling point, a distance from the eyepoint to that feature at the sampling point,

g. feature storing means for storing feature describing data for each sampling point within a sub-area, the stored dam being indicative of at least the distance of the opaque feature which covers the sampling point and is nearest to the eyepoint and the distance and translucency of at least one nearer translucent feature which covers the sampling point,

h. sampling point output means for producing an output for each sampling point within a sub-area, the sampling point output corresponding to the combined effects of the features identified by the data stored in the data storing means,

i. pixel producing output means for producing an output for each pixel within a sub-area, the pixel output corresponding to the combined effects of the sampling point outputs for all sampling points which contribute to that pixel, and

j. display means for displaying the pixel outputs.

2. An apparatus according to claim 1, wherein each sub-area covers a plurality of pixels.

3. An apparatus according to claim 1, comprising a plurality of parallel sampling point processors each including a said sampling point cover determining means, a said distance determining means, a said storing means and a said sampling point output producing means.

4. An apparatus according to claim 3, comprising one sampling point processor for each sampling point within a sub-area

5. An apparatus according to claim 1, comprising a plurality of parallel pixel processors each including a said pixel producing output means.

6. An apparatus according to claim 5, comprising one pixel processor for each pixel within a sub-area.

7. An apparatus according to claim 1, wherein the pixel producing output means receives inputs relative to each pixel from each sampling point lying within the pixel and each sampling point lying within a predetermined area extending around the pixel.

8. An apparatus according to claim 7, wherein the predetermined area is a half pixel wide strip extending around the pixel.

9. An apparatus according to claim 1, wherein the sampling point cover determining means comprises, relative to each sampling point, means for calculating a perpendicular distance for a sampling point from edges of any polygonal feature that at least partially covers a respective sub-area, and means for logically combining the calculated perpendicular distances to determine whether the sampling point is within the polygon.

10. An apparatus according to claim 9, wherein each polygon edge is defined by an edge equation in said model, the coefficients of the edge equations being selected such that the polygon edges form a closed string of connected vectors all running in the same direction around the polygon, the calculating a perpendicular distance means produces an output relative to each edge equation, the sign of said output indicates the side of the edge on which the sampling point is positioned, and the logical combining means comprises means for indicating whether the sampling point is within the polygon by reference to the signs of the distances to the polygon edges.

11. An apparatus according to claim 10, wherein the polygon is convex and its edges form a closed string of connected vectors all running in the clockwise direction around the polygon, the calculating a perpendicular distance means produces a negative output in respect of any one edge equation if the sampling point is to the left hand side of the polygon edge represented by that edge equation, and the logical combining means produces an output indicating that the sampling point is not covered if any one output of the distance calculating means is negative.

12. An apparatus according to claim 1, wherein the sampling point cover determining means comprises, relative to each sampling point, means for calculating a perpendicular distance for a sampling point from major and minor axes of any screen space ellipse feature that at least partially covers a respective sub-area, and means for substituting the calculated distances in a screen equation of the ellipse to determine whether the sampling point is covered by the ellipse feature.

13. An apparatus according to claim 1, wherein the distance determining means comprises means for performing an inverse perspective transformation to calculate from screen space coordinates of the respective sampling point at least one eyepoint space coordinate of a point on the feature which lies on a line drawn from the eyepoint through the sampling point.

14. An apparatus according to claim 13, wherein the said eyepoint space coordinate is the Z coordinate where the Z axis of eyepoint space extends from the eyepoint through a center of the display screen.

15. An apparatus according to claim 1, wherein the feature storing means comprises a translucency stack memory, means for comparing a distance from the eyepoint at the sampling point of each feature received with distances from the eyepoint of any features previously received by the feature storing means, and means for storing each received feature in the stack in a position in the stack such that the features are stored in an order corresponding to the distance from the eyepoint.

16. An apparatus according to claim 15, comprising means for positioning coplanar features in the stack in accordance with feature attribute data defining the relative visibility of coplanar features.

17. An apparatus according to claim 15, wherein the storing means comprises means for storing data identifying the first received opaque feature at the last position in the stack, and means for replacing the data identifying the first received opaque feature with data identifying a subsequently received opaque feature if the subsequently received opaque feature is nearer to the eyepoint than the feature previously stored in the last position in the stack.

18. An apparatus according to claim 17, wherein the storing means comprises means for storing data identifying translucent features in the stack at positions which are sorted in distance from the eyepoint order, translucent features which are further from the eyepoint than any opaque feature identified by data stored in the last position in the stack being discarded.

19. An apparatus according to claim 15, wherein features representing raster light points to be included in the raster pixels of the image have a flag in said model to distinguish them from other features, and the storing means comprises means for limiting the raster light point features which can be stored in the translucency stack.

20. An apparatus according to claim 19, wherein the limiting means is arranged to limit the raster light point features in the translucency stack to one.

21. An apparatus according to claim 1, wherein the sampling point output producing means comprises means for producing red, green and blue outputs derived from the following equations:

space="preserve" listing-type="equation">R=cT.sub.0 ·

R.sub.0 +cT.sub.1 ·

R.sub.1 +. . . cT.sub.N ·

R.sub.N

space="preserve" listing-type="equation">G=cT.sub.0 ·

G.sub.0 +cT.sub.1 ·

G.sub.1 +. . . cT.sub.N ·

G.sub.N

space="preserve" listing-type="equation">B=cT.sub.0 ·

B.sub.0 +cT.sub.1 ·

B.sub.1 +. . . cT.sub.N ·

G.sub.N

where R, G and B are the red, green and blue outputs respectively,

cT_N is a cumulative translucency of a Nth feature in distance from an eyepoint order stored by the feature storing means,

R_N is the red component of the Nth feature in distance from the eyepoint order stored by the storing means,

G_N is the green component of the Nth feature in distance from the eyepoint order stored by the storing means, and

B_N is the blue component of the Nth feature in distance from the eyepoint order stored by the storing means.

22. An apparatus according to claim 1, wherein the pixel producing output means comprises a weighting and accumulation circuit receiving outputs corresponding to each pixel from the sampling point output producing means for each sampling point affecting that pixel, the weighting and accumulation circuit allocating a weighting to each received output of the sampling point output producing means that is a function of the position of each sampling point relative to that pixel, and accumulating the weighted outputs of the sampling point output producing means for producing said pixel output.

23. An apparatus according to claim 1, wherein the sampling point position determining means comprises means for determining corner coordinates in screen space for each sub-area, and means for determining sampling point coordinates by interpolation from the corner coordinates.

24. An apparatus according to claim 1, wherein each sub-area has a boundary in screen space corresponding to a boundary of the screen space area that represents a mapping from eyepoint space to screen space of an area of the display screen covered by one or more pixels corresponding to the sub-area.

25. An apparatus according to claim 1, wherein each sub-area covers at least one pixel and a boundary area extending a predetermined distance therearound.

26. An apparatus according to claim 25, wherein the boundary area is equal in width to one half a pixel width.

27. An apparatus according to claim 1, wherein the sub-area coverage determining means comprises means for detecting opaque features which fully cover the sub-area, and means for discarding features which lie farther away from the eyepoint than the fully covering opaque feature.

28. An apparatus according to claim 27, comprising means for calculating the distance from the eyepoint at each corner of the sub-area of a plane defined by each feature, and means for comparing the calculated distances, and discarding a feature if the calculated distances for that feature are greater than corresponding calculated distances for a nearest to the eyepoint opaque feature.

29. An apparatus according to claim 28, wherein the discarding means comprises a store receiving data identifying features at least partially covering the respective sub-area, the comparing means receiving outputs from the store and providing an input to the store to cause data identifying features lying behind the nearest opaque feature to be cancelled from the store.

30. An apparatus according to claim 1, wherein the sub-area coverage determining means comprises means for calculating a perpendicular distance from each corner of the sub-area to the edges of a bounding box defined in screen space coordinates for each feature to include all that feature within its extent, and means for logically combining the calculated perpendicular distances to provide an output indicative of sub-area coverage.

31. An apparatus according to claim 30, wherein the bounding box is defined by edges forming a closed string of connected vectors all running in the same direction around the box and the combining means produces an outcode in respect of each sub-area corner, the outcode for any one corner being a first outcode if all the bounding box edge distances are of one sign or zero, a second outcode if all the bounding box edge distances are of the opposite sign or some are of the opposite sign and some are zero, and a third outcode if any other combination of bounding box edge distances is produced, the outcodes for the corners being combined by gates to form total outcodes a first of which indicates that the bounding box totally covers the sub-area, a second of which indicates that the bounding box partially covers the sub-area, and the remainder of which indicate that the bounding box is outside the sub-area.

32. An apparatus according to claim 31, comprising means for combining the total outcodes to produce final outcodes, a first final outcode indicating that the sub-area is at least partially covered if the bounding box and polygon edge outcodes are both either the first or the second total outcode, and a second final outcode indicating that the sub-area is totally covered by an opaque feature if the bounding box and polygon edge final outcodes are both the first total outcode and the feature is opaque.

33. An apparatus according to claim 1, wherein the sub-area coverage determining means comprises means for calculating a perpendicular distance from each corner of the sub-area to edges of polygonal features, and means for logically combining the calculated perpendicular distances to provide an output indicative of sub-area coverage.

34. An apparatus according to claim 33, wherein the polygonal features are defined by edges forming a closed string of connected vectors all running in the same direction around the polygon, and the combining means produces an outcode in respect of each sub-area corner, the outcode for any one corner being a first outcode if all the polygon edge distances are of one sign or zero, a second outcode if all the polygon edge distances are of the opposite sign or some are of the opposite sign and some are zero, and a third outcode if any other combination of polygon edge distances is produced, the outcodes for the corners being combined by gates to form total outcodes a first of which indicates that the polygon totally covers the sub-area, a second of which indicates that the polygon partially covers the sub-area, and the remainder of which indicate is that the polygon outside the sub-area.

35. An apparatus according to claim 1, comprising super sub-area dividing means for dividing the image area into an array of super sub-areas each of which covers a plurality of said sub-areas, means for storing data identifying features which at least partially cover each super sub-area, and means for reading out data to the sub-area coverage determining means from the storing means one super sub-area at a time.

36. An apparatus according to claim 35, wherein the sub-area coverage determining means processes each sub-area of the super sub-area separately.

37. An apparatus according to claim 36, wherein the sub-area coverage determining means comprises at least two processors, and means are provided to distribute processing tasks to the processors such that adjacent sub-areas are processed by different processors.

38. An apparatus according to claim 35, comprising means for calculating a perpendicular distance from a reference point within a super sub-area to each edge of a feature, means for calculating a limiting distance from the reference point, the limiting distance being such that if a feature edge is at a distance from the reference point greater than the limiting distance that edge cannot cross the super sub-area, means for comparing the calculated perpendicular distances with the limiting distance, and means for assessing super sub-area coverage by a feature on the basis of a logical combination of the results of the comparison between the calculated perpendicular distances and the limiting distance.

39. An apparatus according to claim 38, wherein each super sub-area is rectangular, and means are provided to calculate a limiting distance for each edge of a feature with respect to each corner of the super sub-area, the limiting distance being a perpendicular distance from the center of the super sub-area to a line drawn parallel to a respective feature edge and passing through a respective super sub-area corner.

40. An image generator for use with an image projection system in which an image is projected onto a display surface and the display surface is viewed from a predetermined eyepoint through an imaginary viewing plane of predetermined area, the image being projected as a series of raster scan lines each made up from a respective row of display pixels, wherein the image generator comprises a model database in which features of the model are described by geometrical attribute data defining the feature with reference to world space and non-geometrical attribute data defining characteristics of the feature, means for determining an eyepoint position in world space from which the model is to be viewed, means for transforming geometrical attribute data from world space to eyepoint space, and means for calculating image data to be displayed on the display surface from the transformed geometrical attribute data and the non-geometrical attribute data, the image data being consistent with an appearance of the model from the eyepoint, wherein the image data calculating means comprises:

a. means for dividing a viewing plane area into an array of sub-areas, each sub-area being defined by four comer coordinates arranged such that a projection of the sub-area onto the display surface from the eyepoint corresponds in shape and area to a portion of the display surface upon which a predetermined respective group of pixels is projected,

b. means for defining a position of at least one sampling point within each sub-area, each sampling point position being defined by reference to comers of the respective sub-area,

c. means for determining from the transformed geometrical feature attributes and the position of each sampling point which of the sampling points is covered by each of the features,

d. attribute storing means for storing for each sampling point non-geometrical attribute data for at least one feature covering that sampling point, and

e. pixel output generating means for generating from the stored attribute data an output to the image projection system in respect of each pixel.

41. An image generator according to claim 40, wherein each sub-area has linear edges.

42. An image generator according to claim 40, wherein each sub-area corresponds in area to a rectangular array of pixels including a plurality of rows and columns of pixels.

43. An image generator according to claim 40, comprising a plurality of sampling points per pixel.

44. An image generator according to claim 40, wherein the attribute storing means comprises means for storing for each sampling point attribute data for one opaque and at least one translucent feature covering that sampling point.

45. An image generator according to claim 40, wherein the pixel output generating means comprises means for accumulating and weighting the stored attribute data from a predetermined group of sampling points distributed near a pixel to produce the output to the image projection system.

46. An apparatus for producing an image on a display screen of a world space model made up from a plurality of features defined in world space, the model including features defined light point and non-light point features and being viewed from an eyepoint defined in world space, wherein the apparatus comprises:

a. means for calculating a finite area in screen space to be occupied by each light point feature,

b. intensity calculating means for calculating the intensity of light point features,

c. means for calculating a translucency for each light point feature visible from the eyepoint, the calculated translucency being a function of the calculated intensity, and

d. output producing means for producing outputs to a display device, said outputs corresponding to said calculated finite area, light intensity and feature translucency.

47. An apparatus according to claim 46, wherein the intensity calculating means calculates intensity of light point features as a function of at least the distance from the eyepoint to the light point.

48. An apparatus according to claim 46, comprising means for determining translucency of non-light point features, means for defining the screen space position of a plurality of sampling points distributed across the display screen, means for determining area of screen space which would be occupied by each feature if that feature alone was viewed from the eyepoint, means for determining for each sampling point which features occupy a position in screen space corresponding to that sampling point, and means for storing attributes for each sampling point including translucency for a nearest to the eyepoint opaque feature and any nearer to the eyepoint translucent features which occupy the position in screen space of that sampling point, the output producing means producing outputs to the display device appropriate to the stored attributes at each sampling point.

49. An apparatus according to claim 48, wherein the storing means comprises a translucency stack, and data describing light points which are to be displayed in a raster made up from a plurality of pixels are stored with data describing non-light point features in the translucency stack in order of their distance from the eyepoint, the output producing means producing an output for each sampling point in accordance with the following equations:

space="preserve" listing-type="equation">R=cT.sub.0 ·

R.sub.0 +cT.sub.1 ·

R.sub.1+. . . cT.sub.N ·

R.sub.N

space="preserve" listing-type="equation">G=cT.sub.0 ·

G.sub.0 +cT.sub.1 ·

G.sub.1+. . . cT.sub.N ·

G.sub.N

space="preserve" listing-type="equation">B=cT.sub.0 ·

B.sub.0 +cT.sub.1 ·

B.sub.1+. . . cT.sub.N ·

B.sub.N

where R, G and B are red, green and blue outputs respectively,

cT_N is the cumulative translucency of the Nth feature in distance from the eyepoint order stored by the storing means,

R_N is the red component of the Nth feature in distance from the eyepoint order stored by the storing means,

G_N is the green component of the Nth feature in distance from the eyepoint order stored by the storing means, and

B_N is the blue component of the Nth feature in distance from the eyepoint order stored by the storing means.

50. An apparatus according to claim 49, comprising means for limiting the light points to be displayed in the raster which can be stored in the translucency stack.

51. An apparatus according to claim 50, comprising means for limiting the light points in the translucency stack to one.

52. An apparatus for producing an image on a display screen of a world space model made up from a plurality of features defined by world space coordinates and viewed from an eyepoint defined in world space, the model including light point features each defined by world space coordinates determining a position of the light point in world space, wherein the apparatus comprises:

a. means for calculating screen space coordinates of each light point,

b. means for calculating a screen space area for each light point as a function of at least distance in world space between the light point and eyepoint,

c. intensity calculating means for calculating an intensity for each light point,

d. means for defining the screen space positions of a plurality of sampling points distributed across the display screen using the calculated light point screen space coordinates,

e. means for determining for each light point which of the sampling points lie within the calculated area of the light point, and

f. means for producing an output to a display device for each light point, said output corresponding to the calculated light point intensity and the particular sampling points which lie within the calculated area of the light point.

53. An apparatus according to claim 52, wherein the intensity calculating means calculates intensity as a function of distance between the eyepoint and light point.

54. An apparatus according to claim 52, wherein at least one of the light point features is to be displayed calligraphically, the apparatus further comprising means at each sampling point for calculating the attenuation of, a calligraphic light point within the area of which that sampling point is located, the attenuation accounting for the presence of any other feature which covers the sampling point and is nearer to the eyepoint than the calligraphic light point, and means for calculating a display intensity for the calligraphic light point as a function of the sum of the attenuations of the calligraphic light point at all the sampling points which are positioned within the area of the calligraphic light point and the light point intrinsic intensity.

55. An apparatus according to claim 54, wherein the intensity calculating means calculates an intrinsic intensity for each light point assuming no occultation of the light point by nearer to the eyepoint features, the apparatus further comprising means for calculating a translucency for each calligraphic light point, the translucency being calculated as a function of the calculated intrinsic intensity, and means for calculating a display intensity for calligraphic light points as a function of the intrinsic intensity attenuated by any nearer to the eyepoint opaque or translucent features.

56. An apparatus according to claim 55, comprising calligraphic light point processing means for processing calligraphic light point features in near to liar from the eyepoint order after all non-calligraphic light points have been processed, the calligraphic light point processing memos comprising attenuation calculating means for calculating for each sampling point the attenuation of calligraphic light points coveting that sampling point as a function of translucency of features that are nearer to the eyepoint than the calligraphic light point, means for accumulating for each calligraphic light point the calculated attenuations for all sampling points covered by that light point, and means for producing for each calligraphic light point a calligraphic light point display intensity output which is a function of the accumulated attenuations, the light point display intensity being stored in a calligraphic light point projection device.

57. An apparatus according to claim 56, comprising means for determining an aperture for a calligraphic light point, the aperture being a sum of sampling points potentially covered by the calligraphic light point, means for determining an attenuation factor for a calligraphic light point, the attenuation factor being a sum of the attenuations of the calligraphic light point at each covered sampling point, and the attenuation at each sampling point being a calculation of the proportion of light emitted by the feature represented as the calligraphic light point which in the modelled world would reach the eyepoint, and means for calculating the display intensity for the calligraphic light point by multiplying the intrinsic intensity of the light point by a fraction calculated as the attenuation factor divided by the aperture.

58. An apparatus according to claim 56, wherein the attenuation calculating means comprises a translucency stack for each sampling point, means for storing in the translucency stack data identifying the nearest opaque and any nearer to the eyepoint non-calligraphic light point features which cover that sampling point, a veil store corresponding to each sampling point, means for comparing the distance from the eyepoint of each calligraphic light point with the distances from the eyepoint of features stored in the translucency stack, and means for storing in the veil store a cumulative translucency for the calligraphic light point, the cumulative translucency being representative of the attenuation of the calligraphic light point by nearer to the eyepoint features stored in the translucency stack.

59. An apparatus according to claim 58, wherein the calculated attenuation and aperture for each potentially impacted sampling point are combined with those of other impacted sampling points and output to the light point store.

60. An apparatus for scan converting data describing a plurality of features to enable the display of an image of a world space model defined by those features, each feature having a boundary defined by a plurality of straight edges, and each edge being defined by a line equation in screen space coordinates, the apparatus comprising means for dividing the screen into a plurality of sub-areas, and means for analyzing the coverage of any one sub-area by any one feature, wherein the coverage analyzing means comprises:

a. means for calculating a perpendicular distance from a reference point within the one sub-area to each edge of the one feature,

b. means for calculating a limiting distance from the reference point, the limiting distance being such that if a feature edge is at a perpendicular distance from the reference point which is greater than the limiting distance that edge cannot cross the sub-area,

c. means for comparing the calculated perpendicular distances with the limiting distance, and

d. means for assessing coverage of the sub-area by a feature on the basis of a logical combination of results of the comparisons between the calculated distances and the limiting distance.

61. An apparatus according to claim 60, wherein each sub-area is rectangular and means are provided to calculate a limiting distance for each edge of a feature with respect to each corner of the sub-area, the limiting distance being the perpendicular distance from a center of the sub-area to a line drawn parallel to a respective feature edge and passing through a respective sub-area corner.

62. An apparatus for processing data describing a model an image of which is to be displayed on a screen, the model comprising a plurality of features each described in terms of geometrical attributes defining position and orientation of the feature and non-geometrical attributes defining characteristics of the feature, and the image being intended to represent a view of the model from an eyepoint in world space, comprising

a. a database in which model data is stored in a hierarchical tree structure having a root node corresponding to a base of the tree, branch nodes corresponding to branching points of the tree, and leaf nodes corresponding to ends of individual branches of the tree, each node of the tree storing data describing a respective object, leaf nodes storing at least one feature of the respective object, and root and branch nodes storing at least one pointer to another object and transformation data defining a relative position and orientation of the pointed to object relative to a pointed from object, whereby successive locations in the tree structure store data related to successively more detailed portions of the model,

b. a transformation processor having a parallel army of object processors, and

c. a controller for reading out data from the database to the object processors such that data read out from one node is read out to a corresponding one of a plurality of data processors,

wherein each data processor is adapted to receive data read out from any node of the tree, one of said object processors responding to read out of pointer data by transforming the respective transformation data into a common coordinate space and returning the pointer and the transformed data to the controller, and the one object processor responding to read out of a feature by transforming the geometrical attributes of the feature, the controller reading out data stored at the node corresponding to the base of the tree and subsequently to read out data stored at nodes of the tree identified by the pointers returned to it from the object processors.

63. An apparatus according to claim 62, wherein the controller reads out object data to a bus to which each of said object processors are linked, means are provided to indicate whether an object processor is idle and available to receive object data for processing, and means are provided for enabling an idle object processor to receive object data made available on the bus, the object processor receiving the object data indicating that the object data has been received for processing to prevent the same object data being received by other object processors.

64. An apparatus according to claim 63, wherein the object processors are arranged in groups, each group communicating with the bus through a local controller, the local controller accepting object data from the bus only if at least one of the respective group of object processors is idle, and the local controller indicating by a signal applied to the bus when it has accepted object data for processing by one of its respective group of object processors.

65. An apparatus according to claim 63, wherein object data is distributed to object processors on a pseudo-random basis.

66. An apparatus according to claim 63, wherein each object processor comprises means for indicating completion of processing of object data read out to that object processor.

67. An apparatus according to claim 62, wherein each of the object processors responds to receipt of pointer and feature data describing an object firstly by processing the pointer data and returning the pointers to the controller and secondly by processing the feature data.