Key Frames Extraction from Human Motion Capture Data Based on Hybrid Particle Swarm Optimization Algorithm
Key Frames Extraction from Human Motion Capture Data Based on Hybrid Particle Swarm Optimization Algorithm
复制标题
基于混合粒子群优化算法的人体动作捕捉数据关键帧提取
DOI：
10.1007/978-3-319-31277-4_29
复制
发表时间：
2016
期刊：
影响因子：
--
通讯作者：
Qiang Zhang
中科院分区：
文献类型：
--
作者：
Xiaojing Chang;Pengfei Yi;Qiang Zhang
从人体运动捕捉数据中提取关键帧是近年来计算机动画研究的热点问题。虽然现有方法对关键帧的重建误差较小，但关键帧的数目仍需减少。为了得到关键帧少、重建误差小的结果，提出了一种基于混合粒子群优化算法的关键帧提取方法。该方法将遗传算法的进化策略引入到混合粒子群优化算法中，能够得到压缩比最优且重构误差小的关键帧。实验结果表明了该方法的有效性。
Extracting key frames from human motion capture data is a hot issue of computer animation in recent years. Though the reconstruction error of the key frames by current methods is small, the number of key frames still needs to be reduced. In order to produce results with less key frames and small reconstruction error, we propose a method employing hybrid particle swarm optimization algorithm to extract key frames. By introducing evolution strategy of Genetic Algorithm (GA) to hybrid particle swarm optimization algorithm, the method can get key frames with optimal compression ratio and small reconstruction error. Experimental results show the effectiveness of our method.