Detection method for collision between solid mesh models based on GPU (Graphics Processing Unit) acceleration

Abstract

The invention discloses a detection method for the collision between solid mesh models based on GPU (Graphics Processing Unit) acceleration. Compared with the conventional method, the detection method is characterized in that aiming at all vertexes in the models, parallel rays are created, and the same ray length is set as the collision precision; the collision precision can be manually adjusted according to the requirement; all collision rays in the collision detecting process are subjected to the same execution procedures, so that intersection calculation for all collision rays and all triangular patches of the models is realized by applying the GPU, the calculation time is shortened, the calculation efficiency is improved, and the detection method for real-time accurate collision between the solid mesh models is realized.

Description

technical field [0001] The invention relates to the field of virtual assembly, in particular to a method for detecting collisions between solid mesh models based on GPU acceleration. Background technique [0002] Virtual assembly is one of the important research directions emerging in recent years. From the perspective of product assembly design, it uses virtual reality technology and computer simulation technology to establish a multi-mode virtual environment with auditory, visual and tactile sensations. With input and output devices, designers can conduct assembly operations and planning interactively in a virtual environment, inspect and evaluate product assembly performance, and generate economical, reasonable, and practical assembly solutions. [0003] Collision detection is an inevitable technical problem in virtual assembly, especially for the assembly design of large-scale complex scientific devices. The real-time and accuracy of collision detection directly affect t...

AI Technical Summary

Problems solved by technology

Based on this idea, there have been some collision detection algorithms that perform partial detection calculations on the GPU, such as the method of triangle intersection calculation on the GPU and the method of bounding box flow calculation on the GPU. However, when the complexity of the model in the scene is high , if all of them are handed over to the GPU for parallel processing, the efficiency will drop significantly, and the purpose of accuracy and real-time cannot be achieved.

Images