A Motion Graph Approach for Interactive 3D Animation using Low-cost Sensors

摘要: Interactive 3D animation of human figures is very common in video games, studios and virtual environments. However, it difficult to produce full body that looks realistic enough be comparable studio quality motion data. The commercial capture systems are expensive not suitable for everyday Real-time requirements tend reduce animation. We present a graph based framework high sequences real-time using set inertial sensor controllers. user’s action generates signals from the controllers provide constraints select appropriate sequence motions structured database motions, namely graph. Our local search algorithm utilizes noise prone rapidly varying input querying large real-time. ability waive producing provides simple user interface intuitive use. proposed low cost easy setup.