手背静脉图像骨架特征提取的算法

Study on algorithm for skeleton features extraction of hand vein image

手背静脉近红外图像识别是一种新的非接触式生物认证技术。对采集的手背静脉图像进行了增强处理。对二值化图像采用四邻域区域生长的方法,去除噪声斑块。对处理后的静脉图像采用了一种快速细化的细化算法。分析和解决了细化后特征点———交叉点的提取。针对细化后骨架中所引入的另一类噪声———毛刺和静脉图像细化后的特点,提出了一种毛刺修复算法。实验结果表明,经过该算法处理后得到的骨架图像,能够较好的反映静脉纹理特性。

Biometric identification technology based on hand vein subcutaneous network structure appears as a promising technique for personal recognition. The contrast of hand vein image is enhanced by using Wiener filter, normalized mode, gray level stretching, sharpening and so on. After binary image was obtained, four-domain region growing algorithm was presented to differentiate and get rid of the noise areas. The quick thinning algorithm was adopted, which had many advantages such as complete thinning, fast speed, maintenance of minutiae, and so on. After hand vein image skeleton was analyzed, the method for extracting the crossing points was proposed to solve the problems met in the feature extraction. The burr, other kind of noise, was intredueed by thinning algorithm. On the basis of the burr of the hand vein image, an algorithm for deburring was presented. The experiments show that the new algorithm could make the skeleton express geometric structure of the hand vein image.