基于历史的动态手势识别  

History-Based Dynamic Hand Gesture Recognition

作　　者：黄国范[1] 程小平[1] 

机构地区：[1]西南大学计算机与信息科学学院,重庆400715

出　　处：《西南大学学报（自然科学版）》2009年第1期106-110,共5页Journal of Southwest University(Natural Science Edition)

摘　　要：基于表观建模的方法是当前手势识别的主流,我们采用基于历史的手势表示方法,用皮肤历史图像建立手势表观模型,然后用原型模板匹配的方法进行手势识别,并使用切线距离来进行测试样本与模板之间的相似性度量,以消除视觉敏感问题.实验结果表明,该方法显著地提高了识别准确率.A method of hand gesture recognition is proposed using appearance-based modeling in this paper. Hand gesture is represented by history-based approach, building a hand model using skin history image. Hand gesture is recognized by prototype template matching. Tangent distance is used for the first time to measure the similarity between the test sample and the template, thus eliminating visual sensitivity. The experiment showed that our method has markedly improved recognition rate.

关 键 词：切线距离 皮肤历史图像 k_means算法 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]