Method for increasing efficiency of continuous collision detection

Abstract

The invention provides a method for increasing efficiency of continuous collision detection. The method comprises the steps of acquiring coordinates of vertexes of two triangles required to be subjected to collision detection in a space, obtaining cubic equations for collision according to vector coplanarity, obtaining a Sturm sequence of a cubic function according to the definition of the Sturm sequence, substituting boundary values of a specified interval into the sequence so as to respectively obtain two symbolic variable values of boundary value sequences, finally, determining the existence circumstances of the equations in specified interval roots by a difference value between the two symbolic variable values according to Sturm's theorem, and filtering out the cubic equations without roots, which indicate that the collision cannot occur. According to the method, the majority of detection pairs, which cannot be subjected to collision, can be effectively filtered out, the filtering ratio can reach 90% to the highest, and the efficiency of a continuous collision detection algorithm can be remarkably increased.

Description

technical field [0001] The invention relates to a method capable of improving the efficiency of continuous collision detection, in particular to a CCD filtering calculation method based on Sturm's theorem. Background technique [0002] In applications such as physics-based simulation, robot motion planning, tactile rendering, and tolerance verification of virtual prototypes, non-penetrating constraints are widely applied to movable or deformable objects to achieve collision results. Continuous collision detection is a primary technique for maintaining non-penetrating constraints and better handling of collision feedback. [0003] The reason why collision detection is required between objects is that two or more impenetrable objects cannot exist in the same space area in the real world. With the rise of virtual reality, computer animation and other technologies, people are eager to simulate the real environment and the imagined environment that does not exist objectively, an...

AI Technical Summary

Problems solved by technology

[0004] Among them, the continuous collision detection algorithm is defined as in a time parameter interval [t 0 ,t 1 ] to detect whether there is any collision between all objects and other objects. The biggest defect is that it is impossible to exclude adjacent pairs of primitives that have not collided with each other, even if it is a very flat deformable triangular mesh without self-collision. It is also impossible to rule out, so many false positives (False Positives) are generated, and the rejection rate is very low, and each false positive needs to perform a point-surface basic test or an edge-edge basic test, that is, a cubic equation needs to be solved, so , so that most of the collision detection time is spent on false positives, resulting in a large amount of calculation and slow detection speed. In some complex simulation environments, the real-time performance of the test cannot be guaranteed.

However, the calculation time of these two algorithms is long, and the filtering algorithm itself has a large calculation cost.

Images