Feature-based spatial hash continuous collision detection method

Abstract

The invention discloses a feature-based spatial hash continuous collision detection method. The method comprises the following steps: acquiring a physical grid model of an object to be detected; acquiring adjacency information of each feature in the physical grid model; establishing a structure constraint relationship of the physical grid model according to the adjacency information; calculating the predicted position of each particle in the physical grid model at the next moment; and detecting feature pairs which are actually collided in the physical grid model based on the spatial hash collision of each feature. Compared with a triangle-based spatial Hash collision detection method, the feature-based spatial Hash continuous collision detection method provided by the invention has the advantages that repeated detection is completely eliminated, and the elimination efficiency is high; compared with other feature-based BVH methods, the method has the advantages that extra time-consuming steps are avoided, the development difficulty is lower, parallel operation is easy, and collision and self-collision detection can be realized without difference; the method is high in collision detection efficiency, low in time consumption and easy to implement.

Description

technical field [0001] The invention belongs to the technical field of continuous collision detection, and in particular relates to a feature-based space hash continuous collision detection method. Background technique [0002] Physical modeling of various objects in the real world (including rigid bodies, soft bodies, and flexible objects) based on physical simulation technology, using computer computing and rendering capabilities to display dynamic and realistic effects has been widely used in the film industry, electronic games, Virtual reality and many other fields. Collision detection is a very important and time-consuming part. At present, there are many methods to speed up collision detection, which can be roughly divided into two categories. One is based on object division, such as the hierarchical bounding box tree (BVH) method, and the other is Based on space division, such as space octree, binary space partition tree (BSP), K-d tree and space hash method, etc. H...

AI Technical Summary

Problems solved by technology

[0004](1) Excluding duplicate collision detection requires assigning each feature to only one adjacent triangle, which takes additional time, especially when mesh topology occurs When changing, the broken part needs to be redistributed;

[0005] (2) The above improved algorithms are based on the hierarchical bounding box tree. For flexible objects such as cloth, when large deformation occurs, the removal efficiency of BVH will be greatly improved. Reduction, the BVH tree needs to be adjusted, and even needs to be rebuilt if necessary, which will seriously affect the performance of collision detection;

[0006](3) The BVH tree method needs to use different traversal methods for the collision detection between objects and the self-collision detection of objects, and also needs to choose some optimization strategy for BVH The tree is adjusted, which increases the difficulty of algorithm implementation

[0007] Relatively speaking, the spatial hashing method can avoid the above points, and can easily realize parallel computing, but as far as the applicant knows, there is no Continuous Collision Detection Method Based on Spatial Hashing of Eigen

Images