摘要
手形识别是基于手形生物特征提取和匹配的人体生物特征识别技术之一。针对传统手形识别技术中的手形矫正方法无法完全排除噪声干扰和识别起始点定位错误的问题，提出一种基于手形凹点和凸点的手形矫正方法的方法。该方法通过提取手性的外侧轮廓，自适应的检测手形中手指的凹凸特征点，克服了背景噪声和手形原始图像中手指识别起始点定位错误。然后利用得到的指形凹凸特征点对手形图像进行位置和方向矫正。
Palm identification is common method using in human body＇s multibiometrics recognition technology based on human palm＇s character extraction and matching. According to the problem that common methods of palm recognition can not clear up noise completely or can not recognize correct points for starting occasionally, the paper advances a method for A palm shape image＇s direction rectification based on points of finger＇s tops and bottoms, which can automatically recognizes points of finger＇s tops and bottoms by extracting palms＇ figure, and which overcomes the mistakes resulted from noise points and false points for starting. Then rectifying the location and direction of palm image by characters of finger＇s tops and bottoms.
作者
郭冉
GUO Ran （Department of Computer Information and Science, Southwest Forestry University, Yunnan 650224, China）
出处
《电脑知识与技术》
2009年第5期3491-3494,共4页
Computer Knowledge and Technology
基金
云南省教育厅科学研究基金（07Y40068）
云南省自然科学基金（2008ZC047M）
关键词
手形方向校正
指形校正
手形轮廓提取
指形凹点
指形凸点
palm shape direction rectification
finger shape recognition
palm figure extracting
finger top points
finger bottom points
