| 基于HMM方法的动态手势轨迹训练性能研究 |
| |
| 引用本文： | 张博洋,吴晓娟,葛庆国,王磊. 基于HMM方法的动态手势轨迹训练性能研究[J]. 信号处理, 2004, 20(6): 662-666 |
| |
| 作者姓名： | 张博洋 吴晓娟 葛庆国 王磊 |
| |
| 作者单位： | 山东大学信息科学与工程学院,济南,250100 |
| |
| 基金项目： | 山东省自然科学基金(Y2001G04)中科院沈阳自动化研究所开放课题基金资助 |
| |
| 摘 要： | 基于HMM(Hidden Markov Model,隐形马尔可夫模型)对动态手势轨迹的训练是手势识别的关键技术之一。本文对HMM的模型训练采用Baum.Welch算法,并分别从迭代次数,样本个数选取,以及模型初值选取等方面对动态手势轨迹的训练性能进行了研究。实验结果表明HMM方法对具有时空特性的动态手势轨迹识别是非常有效的。
|
| 关 键 词： | 手势轨迹识别 HMM Baum-Welch算法 |
| 修稿时间： | 2003-07-10 |
Performance Research of Training Dynamic Gesture Track Based on HMM |
| |
| Zhang Boyang Wu Xiaojuan Ge Qingguo Wang Lei. Performance Research of Training Dynamic Gesture Track Based on HMM[J]. Signal Processing(China), 2004, 20(6): 662-666 |
| |
| Authors: | Zhang Boyang Wu Xiaojuan Ge Qingguo Wang Lei |
| |
| Abstract: | The recognition research based on HMM (Hidden Markov Model) is one of the key techniques in dynamic gesture recognition. This paper adapts the arithmetic of Baum-Wlech to train the HMM and does some researches to the performance of dynamic gesture track form iteration times, selection of the sample number and initial value selected to train the model. The experiment results show the HMM is very efficient to the dynamic gesture track recognition with the space-time characteristic. |
| |
| Keywords: | gesture track recognition HMM training baum-welch arithmetic |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
