摘要
在实时指纹识别系统中,图像增强的效果好坏将影响指纹特征提取及指纹鉴别的识别率。为了解决指纹采集仪采集图像对比度差而且指纹损坏严重情况下识别效果不佳的问题,采用了对反锐化掩膜后指纹图像采用Gabor滤波方法进行图像增强的算法。用反锐化掩膜法预处理指纹图像,并用Sobel算子求取图像的水平和垂直梯度图,把预处理后图像与Gabor函数卷积得到增强图像,可以恢复指纹采集图像中难以识别区域,清楚再现指纹纹理。结果表明,掩膜法与Gabor滤波算法结合在指纹图像增强中可以达到理想的图像增强效果。
In real - time fingerprint recognition system, the quality of the enhanced image is related with feature extraction and fingerprint recognition rate closely. In order to solve the problem of poor recognition owing to the poor contrast and damged fingerprint, a new kind of fingerprint image enhancement algorithm is proposed, which is based on Gabor filter and Counter - peaked mask. The method includes pretreatment by the counter - peaked mask, getting the horizontal and vertical gradient maps by Sobel operator and convoluting with the Gabor function. It can restore the damaged region and reproduce clear texture. The experimental results show that the counter - peaked mask combined with Gabor filter in the fingerprint image enhancement can achieve ideal image enhancement effect .
出处
《计算机仿真》
CSCD
北大核心
2009年第11期220-223,共4页
Computer Simulation
基金
西北工业大学研究生创新实验室中心资助项目(07043)
关键词
反锐化掩膜
加博滤波
指纹增强
圆形滤波
Counter - peaked mask
Gabor filter
Fingerprint enhancement
Circular filter
