手掌静脉识别中感兴趣区域的选择与定位研究

针对目前的手掌静脉识别以掌心区作为感兴趣区域(ROI)受到部分人群掌心区掌脉成像不清晰以致影响整个识别系统性能。这一问题,本文对手掌上的掌心区、大鱼际区和小鱼际区3个区域进行医学分析和3区域静脉吸光量对比实验,选择出大鱼际区作为ROI区域。进而,提出基于大鱼际区的ROI定位方法,以靠近手腕侧最大内切圆确定手掌上的两个稳定特征点,利用这两个稳定特征点确定ROI。在自建的100人的掌脉图库中,对本文方法提取的ROI和传统方法提取的掌心区ROI图像进行掌脉清晰度和掌脉丰富程度比较,实验结果表明,大鱼际区掌脉信息更清晰、丰富。最后,对大鱼际区和掌心区的ROI图像进行识别性能对比实验,结果表明,本文方法使多种识别性能指标都得到明显改善。

Study of ROI selection and location for palm vein recognition

Most of the current palm vein recognition systems use the center of pa lm as the region of interesting(ROI),while troubled with the fuzziness of palm v ein on p alm center area,consequently influencing the performance of the system.To solve this problem,this paper performs medical analysis and contra st experiment of light absorption in the three areas on the palm:the center area of palm,the thenar area and the hypoth enar area.The thenar area is selected as the main scope of ROI.According to this result,a location method is proposed based on the thenar area.The method finds two stable characteristic points and squares the R OI with these two points.In the self-built palm vein database with 100distinct subjects,this study conducts im age definition experiment and abundance experiment on ROI extracted by this method and traditional method firs tly.The result of palm information in the thenar area is clearer and richer than that in the center of the palm.Finally,recognition experiments on the thenar ROI and the center of palm ROI are done.The results s how that the thenar ROI enhances the recognition performance.In conclusion,the thenar area is s uitabl e for ROI and the proposed method enhances the performance of palm vein recognition significantly.