Presentation is loading. Please wait.

Presentation is loading. Please wait.

CS 326A: Motion Planning ai.stanford.edu/~latombe/cs326/2007/index.htm Collision Detection and Distance Computation.

Similar presentations


Presentation on theme: "CS 326A: Motion Planning ai.stanford.edu/~latombe/cs326/2007/index.htm Collision Detection and Distance Computation."— Presentation transcript:

1 CS 326A: Motion Planning ai.stanford.edu/~latombe/cs326/2007/index.htm Collision Detection and Distance Computation

2 Basic problem Given two objects A and B, determine whether they collide (overlap), or not Applications:  Computer graphics  Simulation, e.g., surgical simulation  Robotics, motion planning

3

4 C-Space Sampling (Configurations and Local Paths)  Need for efficient collision-checking algorithms

5

6 Collision Checking vs. Distance Computation Distance is in the workspace between the two closest points distance = 0  collision

7 Two Main Approaches  Feature tracking (pairs of closest features)  Hierarchical bounding volume hierarchies (pre-computation)

8 Feature Tracking Methods  Fixed or arbitrary small discretization This class’s papers: 1. Lin and Canny method (+ V-Clip of B. Mirtich) 2. Application to detecting self-collisions in humanoid robots [Kuffner et al.]  Only update the tracked features at “critical events” when they may change  KDS (Kinetic Data Structure methods) [Guibas et al.]


Download ppt "CS 326A: Motion Planning ai.stanford.edu/~latombe/cs326/2007/index.htm Collision Detection and Distance Computation."

Similar presentations


Ads by Google