摘要
针对指纹图像对比度不强、非线性失真等问题,提出了一种基于模糊聚类和非线性插值的指纹图像增强改进算法。该算法汲取了模糊C-均值聚类和非线性插值算法的优势,在非线性插值算法的基础上使用模糊C-均值聚类分割,确定插值点的坐标和插值公式。试验结果与3种常用的指纹图像算法对比分析表明,新算法对各种因素所致的劣质指纹图像具有明显的增强效果,提高了指纹图像的识别效率。为目前广泛应用的指纹识别系统升级换代提供了一种新的理论和技术支撑。
To solve the problem of the low contrast and nonlinear distortion of fingerprint image, a new image enhancement algo rithm based on fuzzy clustering and nonlinear interpolation is proposed. Our scheme combines the advantage of the fuzzy Cmeans clustering with the advantage of the nonlinear interpolation algorithm. On the basis of the nonlinear interpolation algorithm, the algorithm is designed with the fuzzy Cmeans clustering image segmentation, interpolation points and interpolation formulas. The comparative experimental results show that the new algorithm can obviously enhance the inferior fingerprint image caused by various factors, and improve the efficiency of fingerprint image recognition. This study is significant for both the theoretical research and practical application.
出处
《计算机工程与设计》
CSCD
北大核心
2013年第3期993-997,共5页
Computer Engineering and Design
基金
2012年度河南省科技攻关基金项目(122102210177)
关键词
图像识别
指纹图像
图像增强
模糊聚类
非线性插值
拒识率
image recognition
fingerprint image
image enhancement
fuzzy clustering
nonlinear interpolation
false reject rate
