40
The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL Collision Detection for Deformable Objects Xin Huang [email protected] 16/10/2007

Collision Detection for Deformable Objects

  • Upload
    marinel

  • View
    69

  • Download
    0

Embed Size (px)

DESCRIPTION

Collision Detection for Deformable Objects. Xin Huang [email protected] 16/10/2007. Application. Overview. Deformable models deforming over time, cutting, breaking ….. expensive to update the collision query structure such as BVH as model deforming. Overview. Self- collision - PowerPoint PPT Presentation

Citation preview

Page 1: Collision Detection for Deformable Objects

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

Collision Detection for Deformable Objects

Xin [email protected]/10/2007

Page 2: Collision Detection for Deformable Objects

2

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

Application

Page 3: Collision Detection for Deformable Objects

3

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

Overview

• Deformable models deforming over time, cutting, breaking

….. expensive to update the collision query

structure such as BVH as model deforming

Page 4: Collision Detection for Deformable Objects

4

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

Overview

• Self- collision many adjacent or close primitives

overlap result in a high number of false

positives very challenging

Page 5: Collision Detection for Deformable Objects

5

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

Outline

• [Larsson 01] Collision detection for continuously deforming bodies

• [Zhang 2007] Interactive Collision Detection for Deformable Models

• [Govindaraju 03] CULLIDE: Interactive Collision Detection between Complex Models in Large Environments using Graphics Hardware

• [Govindaraju 05] Interactive Collision Detection between Deformable Models using Chromatic Decomposition

Page 6: Collision Detection for Deformable Objects

6

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

[Larsson 01]

• Collision Detection for Continuously Deforming Bodies [Larsson 01]

hybrid update: an incremental bottom-up and a selective top-down update

Page 7: Collision Detection for Deformable Objects

7

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

[Larsson 01]

• Broad phase: sort and prune• Narrow phase: update, AABBs test,

primitive test

Page 8: Collision Detection for Deformable Objects

8

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

[Larsson 01]

• Bounding volume pre-processing

Built 8-ary AABB tree in top-down manner

A parent AABB is split along three principle axis to form 8 child sub-volumes

Split planes: center point of the box or average of all polygon’s midpoints

Page 9: Collision Detection for Deformable Objects

9

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

[Larsson 01]

• Run-time AABB UpdatesTop-down: traversing the faces under

the node; benefit if a few deep nodesBottom-up: Directly from AABBs of

sub-node; benefit if many deep nodesTrade-off

Page 10: Collision Detection for Deformable Objects

10

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

[Larsson 01]

• Hybrid updateFor a tree with depth n, initially

update the n / 2 first levels bottom-upNon-updated nodes are updated top-

down on the fly during collision traversal

Page 11: Collision Detection for Deformable Objects

11

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

[Larsson 01]

(a) reporting all intersecting triangle pairs (b) first arbitrary intersecting triangle pair

Page 12: Collision Detection for Deformable Objects

12

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

[Larsson 01]

• ConclusionFast to update during deformationNot consider Self-collision

Page 13: Collision Detection for Deformable Objects

13

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

[Zhang 2007]

• Interactive Collision Detection for Deformable Models using Streaming AABBs---[Zhang 2007]

Bound deformable objects as input streams Use GPU to perform parallel pairwise overlap test Compute in object space• Previous GPU based method Depend on image resolution and view direction

Page 14: Collision Detection for Deformable Objects

14

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

[Zhang 2007]

Page 15: Collision Detection for Deformable Objects

15

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

[Zhang 2007]

AABB tree building & texture preparation

Page 16: Collision Detection for Deformable Objects

16

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

[Zhang 2007]

• Streaming AABB Overlap Tests

Page 17: Collision Detection for Deformable Objects

17

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

[Zhang 2007]

• Hierarchical Readback

Page 18: Collision Detection for Deformable Objects

18

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

[Zhang 2007]

• Primitive Intersection Test: on CPU

• Stream Update: on GPU

Page 19: Collision Detection for Deformable Objects

19

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

• Examples

[Zhang 2007]

Page 20: Collision Detection for Deformable Objects

20

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

[Zhang 2007]

• Excluding texture download from CPU to GPU, 2-10 times faster

• Including texture download, 1.4-2 times faster

Page 21: Collision Detection for Deformable Objects

21

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

[Zhang 2007]

Page 22: Collision Detection for Deformable Objects

22

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

[Zhang 2007]

• ConclusionStreaming AABB overlap tests and

stream update using SIMD computations

Stream reduction readbackCollision detection in Object space

Page 23: Collision Detection for Deformable Objects

23

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

[Zhang 2007]

• LimitationPre-setup time to prepare AABB

streams and map to textures in GPU’s memory

Need more texture memoryNot report self-intersections

Page 24: Collision Detection for Deformable Objects

24

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

CULLIDE---[Govindaraju 03]

• CULLIDE: Interactive Collision Detection Between Complex Models in Large Environments using Graphics Hardware

• Compute potentially colliding set (PCS)

• Visibility query by graphics hardware

Page 25: Collision Detection for Deformable Objects

25

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

CULLIDE---[Govindaraju 03]

• An object O does not collide with a set of objects S if O is fully-visible with respect to S.

• Compute PCS by two pass 1st pass, render the objects in the order O1, ..,On 2nd pass, render the objects in the reverse order

On,On−1, ...O1 test if an object is fully visible or not, if not, in

PCS

Page 26: Collision Detection for Deformable Objects

26

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

CULLIDE---[Govindaraju 03]

• Object level pruning perform object level pruning by computing

the PCS of objects.

• Sub-Object Pruning identify potential regions of each object in

PCS

• Exact Collision Detection Triangle-triangle intersection on the CPU

Page 27: Collision Detection for Deformable Objects

27

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

CULLIDE---[Govindaraju 03]

Page 28: Collision Detection for Deformable Objects

28

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

CULLIDE---[Govindaraju 03]

• LimitationNo distance and penetration

information Image-space resolutionCannot handle self-collision

Page 29: Collision Detection for Deformable Objects

29

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

Self-collision

• ChallengeNon-interactive ratesMany adjacent or nearby primitives in

close proximityA high number of false positives

Page 30: Collision Detection for Deformable Objects

30

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

[Govindaraju 05]

• Interactive Collision Detection between Deformable Models using Chromatic Decomposition

• Chromatic Mesh Decomposition• Set-based Self-Collision Detection

Page 31: Collision Detection for Deformable Objects

31

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

[Govindaraju 05]

• Non-Adjacent Collision Detection (NACD): AABB and 2.5D overlap test

• Adjacent Collision Detection (ACD): exact VF and EE elementary tests

Page 32: Collision Detection for Deformable Objects

32

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

[Govindaraju 05]

• Chromatic Mesh Decomposition Independent Sets

Graph Coloring Construct an extended-dual graph G = (V, E) Each primitive pi correspond to a vertex V(pi) in V Add an edge (V(pl), V(pm)) to E if and only if

♦ pl and pm are vertex-adjacent♦ There exists a primitive p in the mesh that both

(pl, p) and (p, pm) are adjacent

Page 33: Collision Detection for Deformable Objects

33

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

[Govindaraju 05]

• AABB Hierarchy Culling Test the Hierarchy against itself Compute non-adjacent primitive colliding

• 2.5D Overlap Tests Extend CULLIDE

Page 34: Collision Detection for Deformable Objects

34

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

[Govindaraju 05]• 2.5D Overlap Tests First pass: Traverse the primitives in Si from

the last to the first. Test if pim is fully-visible

against previously rendered primitives in Sj, i.e. S j

>m

• Second pass: Traverse the primitives in Si from the first to the last. Only test the primitive pi

m which was fully visible in the first pass for potential overlap with the PCS Sj, i.e. Sj

<m

Page 35: Collision Detection for Deformable Objects

35

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

[Govindaraju 05]

• Exact Tests: Non-Adjacent Primitives Merge the PCS of all independent sets Use AABB hierarchy to compute intersecting

pairs Perform EE and VF tests between pairs

• Exact Tests: Adjacent Primitives Check all adjacent primitives for intersection Do not test the shared edge or vertex

Page 36: Collision Detection for Deformable Objects

36

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

[Govindaraju 05]

Page 37: Collision Detection for Deformable Objects

37

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

[Govindaraju 05]

• LimitationMesh with fixed connectivityWork well with a small number of

overlapping pairsChromatic decomposition may

produce a high number of independent sets

Page 38: Collision Detection for Deformable Objects

38

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

Conclusion

• No general or optimal method existed

• Approaches based on BVH have shown to be efficient

• Image-space techniques can achieve highly culling rate, however, is limited by discretization accuracy

• Self-collision still remains challenging

Page 39: Collision Detection for Deformable Objects

39

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

Some other approaches

• Distance Fields• Spatial Subdivision• Stochastic Methods(Refer to [Teschner 2005], a State of the Art

review)

Page 40: Collision Detection for Deformable Objects

40

The UNIVERSITY of NORTH CAROLINA at CHAPEL HILL

References• Survey: LIN M., MANOCHA D.: Collision and proximity queries.

In Handbook of Discrete and Computational Geometry, 2003• Collision detection for deformable objects. Teschner, M.,

Kimmerle, S., Heidelberger, B., Zachmann, G., Raghupathi, L., Fuhrmann, A., Cani, M.-P., Faure, F., Magnenat-Thalmann, N., Strasser, W., and Volino, P. 2005. Computer Graphics Forum

• Larsson T., Akenine-Möller T. 2001. Collision detection for continuously deforming bodies. In Eurographics, pp. 325–333. short presentation.

• Interactive Collision Detection for Deformable Models Using Streaming AABBs, Xinyu Zhang, Young J. Kim, IEEE Trans Visualization & Computer Graphics, 2007

• Govindaraju, N., Redon, S., Lin, M. C., and Manocha, D. 2003. CULLIDE: Interactive Collision Detection between Complex Models in Large Environments using Graphics Hardware. Proc. of Eurographics/SIGGRAPH Workshop on Graphics Hardware

• Interactive Collision Detection between Deformable Models using Chromatic Decomposition, Naga K. Govindaraju, David Knott, Nitin Jain, Ilknur Kabul, Rasmus Tamstorf, Russel Gayle, Ming C. Lin, Dinesh Manocha in ACM SIGGRAPH 2005