手指静脉图像质量评价  

A finger vein image quality assessment method

作　　者：王科俊[1] 刘靖宇[1] 马慧 李雪峰[1] 

机构地区：[1]哈尔滨工程大学自动化学院,黑龙江哈尔滨150001

出　　处：《智能系统学报》2011年第4期324-327,共4页CAAI Transactions on Intelligent Systems

基　　金：国家"863"计划资助项目(2006AA04Z248);国家自然科学基金资助项目(60975022);黑龙江省杰出青年科学基金资助项目(JC200703);哈尔滨市科技创新人才研究专项基金资助项目(2007RFXXG009)

摘　　要：根据手指静脉图像本身的特征,首次提出了手指静脉图像质量评价方法.在分析人类视觉系统(HVS)性能的基础上,将图像有效区域、对比度、位置偏移度、模糊度、信息熵作为评价手指静脉图像质量的参数.并综合分析这些参数,加权获得静脉图像总质量评价函数.实验表明,得到了与主观评价结果相一致的客观评价值.According to the features of finger vein images, a new evaluation method of image quality was proposed in this paper. On the basis of the analysis of the Human Visual System （HVS） function, the image properties such as effective region, image contrast, position offset, fuzzy degree, and information entropy were utilized as the pa- rameters of finger vein image quality assessment. Then an evaluation function of vein image quality was obtained through comprehensive analysis and weighting of those parameters. The experimental results show that the objective evaluation value is in accordance with the subjective evaluation value.

关 键 词：静脉图像 图像质量评价 有效区域 对比度 位置偏移度 模糊度 信息熵 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]