结构分类的字母手势识别算法  

Character Gesture Recognition Algorithm Based on Structure Classification

作　　者：朱笃涛[1] 葛元[1] 王林泉[1] 

机构地区：[1]上海海运学院计算科学与技术系,上海200135

出　　处：《计算机工程》2004年第22期147-148,186,共3页Computer Engineering

基　　金：上海市自然科学基金资助项目(02ZD14053)

摘　　要：提出了基于结构分类的手势识别方法,即在字母手势图像中寻找轮廓所包含的结构特征信息,并以此进行分类判断。这些结构特征中包含了大量的特征信息,如手指个数、手指间距离等,利用这些信息就可以运用传统的结构分类方法对手势加以识别。This paper proposes gesture recognition algorithm based on structure classification, which is to extract structure feature from the contour of alphabet gesture frames and then make the judge by classifying them. These structure feature includes massive information——number of fingers, distances between fingers, etc. By use of these information, the gesture can be recognized through traditional structure-classification method.

关 键 词：模式识别 图像识别 锐化处理 字母手势识别算法 结构分类 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]