Semantic environmental model of the 3D virtual scene

The semantic environmental model of the complex 3D scene based on the geometric model of the scene was presented. The extraction method for semantic information of the sampling points forming the height map in the geometric model of the complex3D scene was proposed. By relations of the obstacles, a bracket-matching-like algorithm and ascertained the path points was designed on which characters could stand and modify the height map. Furthermore, semantic navigation graph of the complex 3D scene was formed, which achieved the extraction of the obstacle semantics and matching of the motion semantics. Height map, modified height map and semantic navigation graph made up of the semantic environmental model of the complex virtual scene, which made full thinking of height, stride and behavior preference of the characters. The model enhances the ability of perceiving the environmental information and distinguishing obstacles automatically for virtual character, realizing automatization of motion matching and variety of behaviors.