naga k. govindaraju, stephane redon, ming c. lin, dinesh manocha

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

Naga K. Govindaraju, Stephane Redon, Ming C. Lin, Dinesh Manocha

University of North Carolina at Chapel Hill

Collision Handling

Contact and collision between objects and between interactive tools with objects Non-trivial to simulate collision, main problems are

Automatically report when collision happensAfter collision detection, objects response in a natural way

Collision Handling

A wide range of techniques are used in collision detection on geometric models

Hierarchical trees formed with bounding boxes or spheres for efficient detect in different levelsPenetration depth should be reported to the handling processThe research issue for computational geometry and algorithms

Problem

Given a large environment with moving objects,

Compute overlapping objectsOverlapping triangles in each pair

Goal

Interactive collision detectionbetween complex objects

Large number of objectsHigh primitive countNon-convex objectsOpen and closed objects

Non-rigid Motion

Deformable objectsChanging topology

Real-time Collision Detection using our Approach

Complex breaking objects on a NVIDIA GeForce FX5800 Ultra GPU

Related WorkObject space techniquesTwo phases

Broad phase – Compute object pairs in close proximity Narrow phase – Check each pair for exact collision detection

Limitations of Object-space Techniques

Considerable pre-processingHard to achieve real-time performance on complex deformable models

Collision Detection using Graphics Hardware

Primitive rasterization – sorting in screen-space

Interference tests

Outline

OverviewPruning AlgorithmImplementation and ResultsConclusions and Future Work

Outline

OverviewPruning AlgorithmImplementation and ResultsConclusions and Future Work

Overview

Potentially Colliding Set (PCS) computationExact collision tests on the PCS

Algorithm

Object LevelPruning

Sub-objectLevel

PruningExact Tests

GPU based PCS computation

Using CPU

Potentially Colliding Set (PCS)

PCS

Potentially Colliding Set (PCS)

PCS

Outline

Problem OverviewOverviewPruning AlgorithmImplementation and ResultsConclusions and Future Work

Algorithm

Object LevelPruning

Sub-object Level

PruningExact Tests

Visibility Computations

Lemma 1: An object O does not collide with a set of objects S if O is fully visible with respect to S

Utilize visibility for PCS computation

set S

Collision Detection using Visibility Computations

Fully VisibleObject O

PCS Pruning

Lemma 2: Given n objectsO1,O2,…,On , an object Oi does notbelong to PCS if it does notcollide with O1,…,Oi-1,Oi+1,…,On

Prune objects that do not collide

PCS Pruning

O1 O2 … Oi-1 Oi Oi+1 … On-1 OnO1 O2 … Oi-1 Oi Oi+1 … On-1 OnO1 O2 … Oi-1 Oi Oi+1 … On-1 On

PCS Pruning

O1 O2 … Oi-1 Oi

PCS Pruning

Oi Oi+1 … On-1 On

PCS Computation

Each object tested against all objects but itselfNaive algorithm is O(n2)Linear time algorithm

Uses two pass rendering approach

PCS Computation: First Pass

O1 O2 … Oi-1 Oi Oi+1 … On-1 On

Render

PCS Computation: First Pass

Fully Visible?

Render

O1

O1 O2 … Oi-1 Oi

PCS Computation: First Pass

Fully Visible?

Render Yes. Does not collide withO1,O2,…,Oi-1

PCS Computation: First Pass

O1 O2 … Oi-1 Oi Oi+1 … On-1 On

Render

Fully Visible?

PCS Computation: Second Pass

O1 O2 … Oi-1 Oi Oi+1 … On-1 On

Render

On

PCS Computation: Second Pass

Render

Fully Visible?

On

PCS Computation: Second Pass

Render

Fully Visible?

Oi Oi+1 … On-1 On

Yes. Does not collide with Oi+1,…,On-1,On

PCS Computation: Second Pass

Render

Fully Visible?

O1 O2 … Oi-1 Oi Oi+1 … On-1 On

Yes. Does not collide with O1,…,On-1,On

PCS Computation

O1 O2 … Oi-1 Oi Oi+1 … On-1 On

Fully VisibleFully Visible

PCS Computation

O1 O2 O3 … Oi-1 Oi Oi+1 … On-2 On-1 On

O1 O3 … Oi-1 Oi+1 … On-1

ExampleO1

O2

O3

O4Scene with 4 objectsO1and O2 collideO3, O4 do not collide

Initial PCS = { O1,O2,O3,O4 }

O3

O1

First Pass

O2

Order of rendering: O1 O4

O3

Fully Visible

O1

Fully Visible

O4O4

Fully Visible

Not Fully Visible

Second Pass

O1

O3

O2O2

Fully Visible

Order of rendering: O4 O1O4O4

Fully Visible

O3

Fully Visible

Not Fully Visible

After two passesO1

O2

O3

O4

Fully Visible

Fully Visible

Potential Colliding SetO1

O2

PCS ={O1,O2}

Algorithm

Object LevelPruning

Sub-object Level

PruningExact Tests

Overlap LocalizationEach object is composed of sub-objectsWe are given n objects O1,…,On

Compute sub-objects of an object Oi that overlap with sub-objects of other objects

Overlap LocalizationOur solution

Test if each sub-object of Oi overlaps with sub-objects of O1,..Oi-1

Test if each sub-object of Oi overlaps with sub-objects of Oi+1,...,On

Linear time algorithmExtend the two pass approach

Potential Colliding SetO1

O2

PCS = {O1,O2}

O1

Sub-objects

O2

PCS = sub-objects of {O1,O2}

First Pass

Rendering order: Sub-objects of O1 O2

Fully VisibleFirst Pass

Fully Visible First Pass

Fully VisibleFully Visible

Not Fully VisibleFirst Pass

Fully VisibleFully Visible

Fully Visible

First Pass

Fully VisibleFully Visible

Fully Visible

First Pass

Fully VisibleFully Visible

Fully Visible

Fully Visible

First Pass

Second Pass

Rendering order: Sub-objects of O2 O1

Fully Visible

Second Pass

Fully Visible

Second Pass

Fully VisibleSecond Pass

Not Fully VisibleFully Visible

Fully Visible

Fully Visible

Second Pass

Fully Visible

Fully Visible

Fully Visible

Fully VisibleSecond Pass

Fully Visible

Fully Visible

Fully Visible

After two passes

PCS

Algorithm

Object LevelPruning

Sub-objectlevel

PruningExact Tests

Exact Overlap tests using CPU

Visibility Queries

We require a queryTests if a primitive is fully visible or not

Current hardware supports occlusion queries

Test if a primitive is visible or not

Visibility QueriesDepth function

GEQUAL LESSAll fragments Pass FailPass

FailFailPassFail PassFail

Query not supported

Occlusion query

Examples - HP_Occlusion_test, NV_occlusion_query

Bandwidth Analysis

Read back only integer identifiers

Independent of screen resolution

Optimizations

Use object bounding volumeUse orthographic views for pruning

Outline

OverviewPruning AlgorithmImplementation and ResultsConclusions and Future Work

Implementation

Dell Precision workstationDell M50 Laptop

Test Models: Environment 1

100 deforming cylindersTotal – 20K triangles

Test Models: Environment 2

Deforming toriiEach torus – 20K triangles

Test Model: Environment 3

250K Dragon35K Bunny

Test Model: Environment 4

Breaking dragon – 250K trianglesBunny – 35K triangles

Implementation

Dell Precision workstation, 2.4GHz Pentium IV CPU, NVIDIA GeForce FX 5800 Ultra GPU, 2GB memory running Windows 2000

Performance

Number of objects v/s collision time

(in m

s)

Performance

Number of polygons v/s collision time

(in m

s)

Performance

Screen resolution v/s collision time

(in m

s)

Advantages

No coherenceNo assumptions on motion of objectsWorks on generic modelsA fast pruning algorithmNo frame-buffer readbacks

Limitations

No distance or penetration depth informationResolution issuesNo self-collisionsPerformance varies with relative configurations

Live Demo of Environment 4

Dell M50 laptop, 2.4GHz Pentium IV-M CPU, NVIDIA Quadro4 700GoGL GPU, 1GB memory running Windows XP

Live Demo of Environment 1

Avg. frames per second = 300

Conclusions

Novel algorithm for collision detection using graphics hardware

No assumptions on motionLinear time PCS computation algorithmNo frame buffer readbacks

Conclusions

First practical solution for general deformable models and breaking objects

Future Work

Handle self-collisionsApply to more interactive applicationsDistance and penetration depth computation