基于局部纹理描述算子的手掌静脉身份识别  

Palm vein identity recognition based on local texture description operator

作　　者：林森[1] 吴微[2] 苑玮琦[3] 

机构地区：[1]辽宁工程技术大学电子与信息工程学院,葫芦岛125105 [2]沈阳大学信息工程学院,沈阳110041 [3]沈阳工业大学视觉检测技术研究所,沈阳110870

出　　处：《仪器仪表学报》2014年第10期2306-2312,共7页Chinese Journal of Scientific Instrument

基　　金：国家自然科学基金(60972123);辽宁省教育厅科学研究一般项目(L2013444;L2014132)资助项目

摘　　要：由于手掌生理结构特性和采集环境因素影响,手掌静脉识别系统中所获取图像容易出现噪声过大、光照不均匀等现象,导致识别性能降低。为了解决这一实际问题,提出一种基于局部纹理描述算子的掌脉识别方法。首先使用Gamma校正、DoG滤波、对比度均衡等手段对掌脉图像进行预处理,降低光照影响。然后将掌脉图像分成互不重叠的区块,使用局部三值模式(LTP)提取每一小块的纹理特征,并加以融合,最后利用卡方距离进行匹配识别。在香港理工大学接触式掌脉图库和自建的非接触式掌脉图库上进行实验测试,结果表明,当图像分块大小为16×16像素时,本方法获得了最高的正确识别率,分别为99.981 6%,99.299 4%,优于其他典型方法,提升了掌脉识别系统的性能,增强了系统的鲁棒性,具有实际应用价值。Due to the influence of physiological structure features of palm and environmental factors, the acquired images in palm vein recognition system are prone to get excessive noise and non-uniform illumination, which may lead to low recognition performance. In or- der to solve this practical problem, a palm vein recognition method based on local texture description operator is proposed. Firstly, the Gamma correction, difference of Gaussian （DoG） filtering and contrast balance methods are used to decrease the influence of illumina- tion on palm vein image preprocessing. Secondly, the palm vein image is divided into some non-overlapping blocks, the local ternary pattern （LTP） is used to extract the texture feature of each block, and then these features are fused. Finally, the chi-square distance is used for matching identification. The experiment tests were conducted on the Hong Kong Polytechnic University （PolyU） contact palm vein image database and the self-built non-contact palm vein image database. Experiment results show that when the image block size is 16 x 16 pixels, the proposed method can obtain the highest correct recognition rates of 99. 981 6% and 99. 299 4% for the two databas- es, respectively. These results demonstrate that this method enhances the recognition performance and robustness of the palm vein identi- fication system, is better than other typical methods, and has practical application value.

关 键 词：掌脉识别 纹理特征 预处理 局部三值模式 

分 类 号：TP391[自动化与计算机技术—计算机应用技术] TH79[自动化与计算机技术—计算机科学与技术]