Fast Adaptive Character Animation Synthesis Algorithm Based on Depth Image Sequence

Abstract

With the continuous development of computer hardware and computer graphics algorithms, computer animation is becoming a hot spot in current research. In this field, character animation is an important research direction. Character animation is the simulation of characters in the real world. The only way to make character animation realistic and make people have a sense of substitution is to ensure the authenticity of characters. In this paper, the motion capture data is constructed into a motion graph, and the problem of generating target motion is transformed into the equivalent problem of searching a path in the graph. At the same time, a cellular based finite state machine is designed to synthesize character reactive motion, creatively combine complex network with crowd animation, and use random graph and uniform network based on cellular structure to synthesize large-scale crowd animation. The multi view expression set is generated semi automatically by curve optimization deformation algorithm, and then the expression space model based on emotion and view elements is established. On this basis, the cartoon expression animation with specific emotion and viewpoint related characteristics can be effectively generated according to the newly input viewpoint and emotion state.