摘要
为了提高掌形认证系统的认证准确率和鲁棒性,提出并实现了一种基于遗传算法的掌形认证方法。该方法首先利用图像采集装置获取掌形图像,然后通过预处理掌形图像提取掌形轮廓特征点集,最后应用遗传算法快速地匹配特征点集完成掌形认证。基于该方法开发了一个原型系统,实验结果表明,该系统的认证准确率高达93%,可应用于各种需要身份认证的场合。
To improve the accuracy and robustness of the hand geometry verification system,a novel hand geometry verification approach based on genetic algorithm is proposed and implemented.Firstly,the hand image is captured by image acquisition device.Then,the image is preprocessed and the geometrical feature point of the hand geometry is extracted.Finally,genetic algorithm is applied to verify the identity of different people by marching the feature points.Based on above approach,a prototype is implemented and experiments results show rate of success in verification up to 93 percent,and the possibility of using this approach in high security environment is verified.
出处
《计算机工程与设计》
CSCD
北大核心
2010年第10期2297-2299,2311,共4页
Computer Engineering and Design
关键词
掌形认证
遗传算法
特征提取
染色体
交叉
变异
hand geometry verification
genetic algorithm
feature extraction
chromosome
crossover
mutation
