关于图像场能分析的指纹图像增强.pdf

第 37 卷第 8 期 光电工程 Vol.37, No.8 2010 年8 月 Opto-Electronic Engineering August, 2010 文章编号：1003-501X(2010)08-0111-06 基于图像场能分析的指纹图像增强 王晓颖，杨 帆，杜颜颜，高 志 ( 河北工业大学 信息工程学院，天津 300401 ) 摘要：手指的脱皮老化以及活体指纹传感器带来的图像误差易影响指纹图像的成像质量。为此，针对低质量指纹 图像中存在的纹线断裂、脊谷粘连以及手指脱皮造成的区域性模糊等问题，提出了一种基于指纹图像场能分析的 指纹图像增强算法。算法分析和计算了指纹图像的强度场、梯度场、方向场和频率场，并在分析图像本身特点的 基础上结合各场能具体信息以及 Gabor 滤波原理对指纹图像进行逐步的增强处理。实践表明，整套算法对低质量 指纹图像增强效果良好，能大幅度提高指纹图像质量。 关键词：指纹图像场；梯度场；方向场；频率场；Gabor 滤波 中图分类号：TP391 文献标志码：A doi：10.3969/j.issn.1003-501X.2010.08.020 Fingerprint Image Enhancement Based on Image Field Energy WANG Xiao-ying ，YANG Fan ，DU Yan-yan ，GAO Zhi ( College of Information Engineering, Hebei University of Technology, Tianjin 300401, China ) Abstract : Finger decrustation, ageing and the image error caused by living fingerprint sensor may affect the imaging quality of fingerprint easily. In order to solve the problems which exist in the inferior quality fingerprint image, such as line breaks, the adhesion of ridge and valley as well as the regional blur caused by finger decrustation, a method based on the analysis of fingerprint image energy is proposed. The algorithm has analyzed and calculated the intensity field, gradient field, directions field and the frequency field of fingerprint image, and then unified the concrete information as well as the principle of Gabor filter to enhance the fingerprint image progressively. The practice indicates that the whole set of algorithm has better enhanced performance and can improve the image quality distinctly. Key words : fingerprint image field; gradient field; direction field; frequency field