一种新的掌纹ROI图像定位方法

在分析传统Harris算法缺陷的基础上,提出一种基于角点检测的快速掌纹图像预处理方法。该方法利用一种改进的自适应Harris算法提取出手掌边缘和手指凹凸处的轮廓特征点,采用属于同一特征组的特征点均值作为候选角点,剔除了邻近角点,并根据候选角点在轮廓线上角度变化的大小定位手指根部的三个关键角点,从而有效提取出掌纹图像的感兴趣区域(ROI)。实验测试图像采用CASIA数据库,实验结果表明,该方法能快速有效地定位出掌纹的ROI区域,且能有效增强掌纹图像的边缘与细节,有利于提高掌纹识别的识别率。

New location method of palmprint ROI images

On the basis of analyzing the disadvantages of traditional Harris algorithm,a rapid palmprint preprocessing method based on angle point detection is proposed in this paper.This method utilizes a modified adaptive Harris algorithm to extract contour features existed in the palm′s edge and the fingers concavo-convex.The feature mean,calculated by features belong to the same feature matrix,is used as the alternative angle-points to delete those adjacent angle points.Then,three key angle-points existed in the fingers′ concavo are located according to the angle change of each angle point in the contour line.Thus,the region of interesting(ROI)is extracted efficiently.In test,the CASIA database is used.Experimental results show that this method can efficiently and quickly locate the ROI of palmprint images,and efficiently enhance the edge and detail of palmprint images simultaneity.