一种基于FSMM的手势识别方法  

A Gesture Recognition Method Based on FSMM

作　　者：邹伟[1] 原魁[1] 杜清秀[1] 陈晋龙[1] 

机构地区：[1]中国科学院自动化研究所高技术创新中心,北京100080

出　　处：《计算机工程》2003年第15期134-135,168,共3页Computer Engineering

基　　金：国家"863"计划资助项目(863-306-zz01-99-2)

摘　　要：结合汉语手语中手势的特点，采用人手的三维空间位置信息作为观测向量，将模糊理论与HMM理论相结合，提出了一种基于FSMM的中国手语手势识别方法。利用观测向量对于各状态的隶属度对Baum-Welch算法进行了改进，实现了FSMM学习递推算法和识别算法。通过实验，证实了该方法的有效性。In this paper, a novel recognition method of Chinese sign language gesture based on the FSMM is introduced, it combines the fuzzy theory with HMM theory. In this method, the 3D position of hand is adopted as the observable vector. The algorithm of Baum-Welch is improved by use of the membership function of every state. The recursive learning algorithm of FSMM and recognition algorithm are implemented. The method is confirmed to be effective by experiments.

关 键 词：FSMM HMM 模糊 手势识别 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]