基于掌纹识别技术的BESTPAL系统  

BESTPAL System Based on Palm Print Identification Technology

作　　者：徐日[1] 张裕[1] 陈晶华 

机构地区：[1]北京电子科技学院,北京100070 [2]上海机电工程研究所,上海201109

出　　处：《无线电工程》2015年第10期29-33,共5页Radio Engineering

基　　金：北京市自然科学基金资助项目(4132066)

摘　　要：传统的掌纹识别系统大多以接触式为主,且设备昂贵、灵活性和移动性较差。为了拓展掌纹识别技术的应用,设计以网络摄像头和手机摄像头为采集端的非接触式掌纹识别系统。获取掌纹图像时,在摄像头的预览画面加限制轮廓,解决在非接触式图像采集过程中因位置不固定而造成掌纹特征纹理图像产生偏移的问题。在掌纹预处理阶段,针对获取的掌纹图像特点,实现一种能够精确定位掌纹特征分离点和截取感兴趣区域的预处理算法,并通过小波变换实现掌纹图像的特征提取和特征匹配。测试结果表明,通过对400人不同掌纹图像的识别处理,BESTPAL掌纹识别系统达到较高的识别率。The traditional palm print identification system are mostly of contact-type with expensive equipment, weak flexibility and mobility.In order to extend the application of palm print identification technology, the non-contact palm print identification system is designed,which uses webcams and mobile phone cameras to get images.When getting the images,by adding restrictions outlines to the image preview of cameras, the system can solve the problem of deviation of palm print feature texture image caused by the unstable palm positions in the collection of palm images.On the pre-processing stage,based on the collected features of palm print images,the system can realize a preprocessing algorithm,which can fix the palm print feature separation point accurately and intercept the interested area of palm print images.In addition,it can implement the feature extraction and matching by wavelet. The test result shows that BESTPAL palm print identification system has higher recognition by testing different palm images of 400 people.

关 键 词：实时掌纹识别 多分辨率分析 小波变换 摄像头 

分 类 号：TP309[自动化与计算机技术—计算机系统结构]