Computer Graphics. Bing-Yu Chen National Taiwan University

Transcription

1 Computer Graphics Bing-Yu Chen National Taiwan University

2 Visible-Surface Determination Back-Face Culling The Depth-Sort Algorithm Binary Space-Partitioning Trees The z-buffer Algorithm Scan-Line Algorithm Visible-Surface Ray Tracing (Ray Casting) Warnock s Algorithm

3 Visible-Surface Determination = Hidden Surface Removal Determining what to render at each pixel. A point is visible if there exists a direct line-of-sight to it, unobstructed by another other objects (visible surface determination). Moreover, some objects may be invisible because there are behind the camera, outside of the field-ofview, too far away (clipping) or back faced (backface culling).

4 Hidden Surfaces: why care? Occlusion: Closer (opaque) objects along same viewing ray obscure more distant ones. Reasons for removal Efficiency: As with clipping, avoid wasting work on invisible objects. Correctness: The image will look wrong if we don t model occlusion properly.

5 Back-Face Culling = Front Facing B x A D C H E F z G

6 Back-Face Culling = Front Facing use cross-product to get the normal of the face (not the actual normal) use inner-product to check the facing N v 1 N = ( v v ) ( v v ) v 2 V v 3

7 Clipping (View Frustum Culling) view frustum eye back face occlusion view frustum

8 List-Priority Algorithms The Painter s Algorithm The Depth-Sort Algorithm Binary Space-Partitioning Trees

9 The Painter s Algorithm Draw primitives from back to front need for depth comparisons.

10 The Painter s Algorithm for the planes with constant z not for real 3D, just for 2½D sort all polygons according to the smallest (farthest) z coordinate of each scan convert each polygon in ascending order of smallest z coordinate (i.e., back to front)

11 The Depth-Sort Algorithm sort all polygons according to the smallest (farthest) z coordinate of each resolve any ambiguities that sorting may cause when the polygons z extents overlap, splitting polygons if necessary scan convert each polygon in ascending order of smallest z coordinate (i.e., back to front)

12 Overlap Cases x y y R P Q Q P P Q z x x

13 Overlap Detection Do the polygons x not overlap? Do the polygons y not overlap? Is P entirely on the opposite side of Q s plane from the viewpoint? Is Q entirely on the same side of P s plane as the viewpoint? Do the projections of the polygons onto the (x,y) plane not overlap?

14 Binary Space-Partitioning Trees An improved painter s algorithm Key observation: f(p)<0 f(p)>0 T 3 T 1 T 5 T 2 T 4 Ax+By+Cz+D=0 f(p): n(p-a)=0

15 Binary Space-Partitioning Trees - T 1 + T 1 T 2 T 3 T 2 T 3

16 Binary Space-Partitioning Trees - T 1 + T 1 - T 2 + T 3 T 2 T 3

17 Splitting triangles a a A c A c B B b b

18 BSP Tree Construction BSPtree makebsp(l: list of polygons) { if (L is empty) { return the empty tree; } Choose a polygon P from L to serve as root; Split all polygons in L according to P return new TreeNode ( P, makebsp(polygons on negative side of P), makebsp(polygons on positive side of P)) } Splitting polygons is expensive! It helps to choose P wisely at each step. Example: choose five candidates, keep the one that splits the fewest polygons.

19 BSP Tree Display void showbsp(v: Viewer, T: BSPtree) { if (T is empty) return; P = root of T; if (viewer is in front of P) { showbsp(back subtree of T); draw P; showbsp(front subtree of T); } else { showbsp(front subtree of T); draw P; showbsp(back subtree of T); } } 2D BSP demo

20 Binary Space-Partitioning Trees P1 P1 A P2 front back 1 P2 P2 D 3 C 2 B back front front back D C A B 3,1,2 3,2,1 1,2,3 2,1,3 extremely efficient for static objects

21 Binary Space-Partitioning Trees Same BSP tree can be used for any eye position, constructed only once if the scene if static. It does not matter whether the tree is balanced. However, splitting triangles is expensive and try to avoid it by picking up different partition planes.

22 BSP Tree

23 BSP Tree inside ones outside ones 2 4 3

24 BSP Tree

25 BSP Tree a 11a 9 9b 11b a 10 11a 5 8 9b 11b 4 3

26 BSP Tree b a 11a 11 11b b 8 4 9a 11b a

27 BSP Tree b a 11a 11 11b b point 10 9a 11b 11a

28 BSP Tree Traversal b a 11a 11 11b b point 10 9a 11b 11a

29 BSP Tree Traversal b a 11a 11 11b b point 10 9a 11b 11a

30 The z-buffer Algorithm Resolve depths at the pixel level Idea: add Z to frame buffer, when a pixel is drawn, check whether it is closer than what s already in the frame buffer

31 The z-buffer Algorithm + = + =

32 The z-buffer Algorithm void zbuffer() { int pz; for (each polygon) { for (each pixel in polygon s projection) { pz=polygon s z-value at (x,y); if (pz>=readz(x,y)) { WriteZ(x,y,pz); WritePixel(x,y,color); } } } }

33 The z-buffer Algorithm y y 1 y 2 y 3 y s z 1 z 2 z 3 z a z p z b Scan line a b p b a b b p s b s a x x x x z z z z y y y y z z z z y y y y z z z z = = = ) ( ) ( ) (

34 z-buffer: Example color buffer depth buffer

35 The z-buffer Algorithm Benefits Easy to implement Works for any geometric primitive Parallel operation in hardware independent of order of polygon drawn Limitations Memory required for depth buffer Quantization and aliasing artifacts Overfill Transparency does not work well

36 Scan-Line Algorithm y E γ+2 γ+1 γ D B β C α F A x

37 Scan-Line Algorithm AET contents Scan line Entries ET entry x y max Δx ID α AB AC β AB AC FD FE γ,γ+1 AB DE CB FE PT entry ID Plane eq. Shading info In-out γ+2 AB CB DE FE ET = edge table PT = polygon table AET = active-edge table

38 General Scan-Line Algorithm add surfaces to polygon table (PT); initialize active-edge table (AET); for (each scan line) { update AET; } for (each pixel on scan line) { determine surfaces in AET that project to pixel; find closest such surface; determine closest surface s shade at pixel; }

39 Ray Tracing = Ray Casting select center of projection and window on viewplane; for (each scan line in image) { for (each pixel in scan line) { determine ray from center of projection through pixel; for (each object in scene) { if (object is intersected and is closest considered thus far) record intersection and object name; } set pixel s color to that at closest object intersection; } }

40 Ray Casting Center of projection Window

41 Ray Casting (Appel, 1968)

42 Ray Casting (Appel, 1968)

43 Ray Casting (Appel, 1968)

44 Ray Casting (Appel, 1968) k a I a + nls i= 1 I i ( ( ) + ( ) ) n k L N k R V d i s i

45 Ray Casting (Appel, 1968) direct illumination

46 Spatial Partitioning

47 Spatial Partitioning R C A B

48 Spatial Partitioning 1 A B 3 2

49 Space Subdivision Approaches Uniform grid K-d tree

50 Space Subdivision Approaches Quadtree (2D) Octree (3D) BSP tree

51 Uniform Grid

52 Uniform Grid Preprocess scene 1. Find bounding box

53 Uniform Grid Preprocess scene 1. Find bounding box 2. Determine grid resolution

54 Uniform Grid Preprocess scene 1. Find bounding box 2. Determine grid resolution 3. Place object in cell if its bounding box overlaps the cell

55 Uniform Grid Preprocess scene 1. Find bounding box 2. Determine grid resolution 3. Place object in cell if its bounding box overlaps the cell 4. Check that object overlaps cell (expensive!)

56 Uniform Grid Traversal Preprocess scene Traverse grid 3D line = 3D-DDA

57 From Uniform Grid to Quadtree

58 Quadtree (Octrees) subdivide the space adaptively

59 Quadtree Data Structure Quadrant Numbering

60 Quadtree Data Structure Quadrant Numbering

61 Quadtree Data Structure Quadrant Numbering

62 Quadtree Data Structure Quadrant Numbering

63 From Quadtree to Octree y x z

64 K-d Tree A A Leaf nodes correspond to unique regions in space

65 K-d Tree A B A Leaf nodes correspond to unique regions in space

66 K-d Tree A B B A Leaf nodes correspond to unique regions in space

67 K-d Tree C A B B A

68 K-d Tree C A B B C A

69 K-d Tree C A D B B C A

70 K-d Tree C A D B B A D C

71 K-d Tree A D B B C C D A Leaf nodes correspond to unique regions in space

72 K-d Tree Traversal A D B B C C D A Leaf nodes correspond to unique regions in space

73 Warnock s Algorithm an area-subdivision algorithm

74 Warnock s Algorithm 1. all the polygons are disjoint from the area 2. there is only one intersecting or only one contained polygon 3. there is a single surrounding polygon, but no intersecting or contained polygons 4. more than one polygon is intersecting, contained in, or surrounding the area, but one is a surrounding polygon that is in front of all the other polygons

75 Warnock s Algorithm surrounding intersecting contained disjoint

76 Warnock s Algorithm x x Contained polygon Intersecting polygon z Intersecting polygon Surrounding polygon z Surrounding polygon

77 Performance of Four Algorithms for Visible-Surface Determination Algorithm Number of Polygons 100 2,500 60,000 Depth sort z-buffer Scan line Warnock area subdivision