基于光谱融合的手掌异常纹识别!.pdfVIP

第35卷摇 第3期 吉林 大学学报 ( 息科 学版) Vol.35摇 No.3 2017年5月 Journal ofJilinUniversity (Information ScienceEdition) May2017 文章编号:1671鄄5896(2017)03鄄0280鄄08 基于光谱融合的手掌异常纹识别 刘摇 闯,刘摇 富,康摇 冰,代立波 (吉林大学 通信工程学院,长春 130022) 摘要:针对现有掌部封闭型病理纹识别算法提取的线特征较少、识别率较低的问题,提出一种基于非下采样剪 切波变换(NSST:Nonsubsample Shearlet Transform)域光谱融合的手掌异常纹识别算法。 首先,选取融合效果最 佳的多光谱掌纹波段组合,并在NSST域内进行多尺度、多方向的分解;其次,根据分解各层子带图像的特点 设计融合规则进行相应系数矩阵的融合,再通过NSST逆变换和形态学处理提取精细纹路特征;然后,利用像 素点的度特点寻找符合要求的闭合纹线回路;最后,采用一种基于矩形度和偏心率等形状描述符的方法识别封 闭型异常纹。 实验结果表明,该识别方法能提取丰富的掌纹线特征,同时,还可准确识别6种不同类型的封闭 型病理纹,识别率可达90%以上。 关键词:非下采样shearlet变换;光谱融合;掌纹;闭合回路;异常纹 中图分类号:TP391.4 文献标识码:A Abnormal Palmprint Recognition Based on Spectral Fusion LIU Chuang,LIU Fu,KANG Bing,DAI Libo (College of Communications Engineering,Jilin University,Changchun 130022,China) Abstract:In order to solve the shortcoming of fewer extracted line feature and lower recognition rates of pathologic palmprint recognition algorithm,weproposed arecognition algorithm of pathologicpalmprint based on spectral fusion in non鄄subsampled shearlet domain. Firstly,the best spectral fused combination of multispectral palmprint is selected. And it is decomposed to the multi鄄directions, multi鄄scales in non鄄subsampled shearlet domain. Next,according to all levels characteristics of sub鄄bands images which had been decomposed,a new fusion rule is designed to fuse the corresponding coefficient matrices. The fine lines feature of palmprint can be obtained by the inverse transformation of the NSST(Nonsubsample Shearlet Transform) and the process of mathematical morphology. Thenthe satisfactory closed circuitsare searchedby degreefeature of pixels. Finally, we proposed a method of combiningthe shape descriptorsbas