基于Harris角点能量的指纹图像分割算法  

Fingerprint segmentation based on Harris corner strength

作　　者：刘花香[1] 王玲[2] 

机构地区：[1]湖南师范大学物理与信息科学学院,长沙410081 [2]湖南大学电气与信息工程学院,长沙410082

出　　处：《计算机工程与应用》2011年第15期206-208,218,共4页Computer Engineering and Applications

基　　金：湖南省教育厅资助科研项目(No.06C522)

摘　　要：指纹图像分割是自动指纹识别系统预处理过程中的一个重要环节。提出了一种基于Harris角点检测器的指纹图像自适应分割算法。首先利用Harris算子计算待分割图像的角点能量,然后使用一种自适应的方法设定阈值完成初分割;采用形态学方法进行后处理以减少分割错误,最后通过面积参数去除分离的残留噪声区域。在FVC2004数据库上的实验结果表明,该分割算法能够准确提取指纹有效区域,并且对低质量指纹图像具有很好的鲁棒性。Fingerprint segmentation is one of critical pre-processing steps in automatic fingerprint identification system.This paper proposes an adaptive fingerprint segmentation based on Harris corner detector.At first, Harris operator is used to compute the strength of Harris corners in the fingerprint image, and then a threshold-setting method is adopted to compute adap- tively the threshold for the early segmentation.Morphology is applied as post-processing to reduce the number of classification errors,the area parameter is used to remove remaining noised region to get the final segmented image.The experimental results on databases of FVC2004 prove that the proposed algorithm provides very accurate segmentation of valid ridge region.Moreover,this classification method also has a great robustness to those fingerprint images with low quality.

关 键 词：Harris角点检测器 指纹图像分割 形态学 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]