基于改进谷形检测增强和Canny边缘修正的手指静脉识别方法  

Finger Vein Recognition Method Based on Improved Valley Detection Enhancement and Canny Edge Correction

作　　者：胡慧鹏 高小明[1] 彭勇[1] HU Huipeng;GAO Xiaoming;PENG Yong(School of Computer Science and Technology, Southwest University of Science and Technology,Mianyang 621010, Sichuan, China)

机构地区：[1]西南科技大学计算机科学与技术学院

出　　处：《西南科技大学学报》2019年第2期92-97,共6页Journal of Southwest University of Science and Technology

摘　　要：手指静脉识别作为新型的生物特征识别技术,未来可广泛应用于身份识别领域。针对当前手指静脉图像质量差导致特征识别准确率较低的问题,提出一种基于改进谷形检测增强和Canny边缘修正的手指静脉识别方法。该方法首先改进用于局部静脉特征增强的方向谷形检测算子,增强静脉特征区域,然后利用Canny边缘检测算法对图像分割后的特征区域进行二次修正,在保证手指静脉特征区域完整性的同时降低伪静脉对识别率的干扰。在手指静脉数据库上进行实验,结果表明该方法对增强算法的泛用性较好,对4种不同的全局图像增强算法的识别准确率分别提升了8.66%,3.7%,1.78%,1.2%,能明显提高特征识别的准确率。As a new biometric recognition technology, finger vein recognition can be widely used in the field of identity recognition in the future.Aimed at the problem that the current finger vein image quality is poor, resulting in low feature recognition accuracy, a finger vein recognition method based on improved valley detection enhancement and Canny edge correction is proposed.The method firstly improves the direction valley detection operator for local vein feature enhancement, enhances the vein feature region, and then uses the Canny edge detection algorithm to perform quadratic correction on the image segmentation feature region to ensure the integrity of the finger vein feature region.At the same time, the interference of the pseudo-vein to the recognition rate is reduced. Experiments on the finger vein database show that the proposed method is more versatile for the enhancement algorithm, and the recognition accuracy of the four different global image enhancement algorithms is increased by 8.66%, 3.7%, 1.78%, and 1.2%,which can significantly improve the accuracy of feature recognition.

关 键 词：图像分割 手指静脉识别 生物识别 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]