中文摘要：

为更好地解决手异质背静脉图像识别问题,本文综合分析了图像采集过程及手背图像的自身特性.图像的有效区域、旋转、位置偏移、对比度、模糊度是造成异质的重要参数.为解决异质手背图像识别问题,提升同类样本的结构相似度,对上述5个质量参数进行优化调整,然后通过实验找出适合不同算法的最佳感兴趣区域（ROI）尺寸.实验结果表明,该方法可以有效提升不同算法对异质手背图像的识别率.

英文摘要：

To better address the problem of the heterogeneity dorsal hand vein recognition,a comprehensive analysis was made of the images acquisition process and the characteristics of hand image.The effective area of the image,the rotation angle,the geometric deviation,the contrast,the blurness are the important parameters that cause heterogeneous images.To solve the problem of heterogeneous dorsal hand images recognition and enhance the structure similarity of similar samples,the five quality parameters were optimally adjusted.Then the experiments found out the best size of ROI（region of interest）for different recognition algorithms.The experimental results show that this method can effectively promote the recognition rate of different algorithms for heterogeneous dorsal hand vein images.