【作者】 王丹；

【导师】 张小波；

【作者基本信息】 西安石油大学 ， 计算机技术（专业学位）， 2020， 硕士

【摘要】 在当前所有的生物识别技术中,指纹识别技术是其中使用最为普遍的技术之一。指纹识别系统的性能在很大程度上依赖于输入指纹图像的质量。通常情况下,受外界环境等因素的影响,指纹图像在获取中,产生的噪声会损坏指纹信息,降低图像对比度,因此需要增强退化的指纹图像以便于后续步骤的处理。本文研究基于各向异性扩散的指纹图像增强方法,主要工作如下:1、对指纹增强的相关理论进行梳理,并阐述指纹图像增强的重要性;2、介绍了基于偏微分方程的线性扩散滤波过程和非线性扩散中的PM、Catte和FABD等模型,以及基于各向异性扩散的边缘增强和相干增强扩散算法,同时对冲击滤波算法进行了探讨;3、针对相干增强扩散滤波增强后的指纹图像会出现边缘模糊及对比度降低等现象,提出一种新的增强模型:CED-SF模型。该模型在第一阶段利用相干增强扩散算法对退化的指纹图像进行修复,之后通过冲击滤波锐化边缘;4、为验证本文方法的有效性,利用MATLAB进行仿真实验,并总结该方法在图像增强等三个方面的应用。先将本文算法、相干增强扩散和冲击滤波这三种方法对指纹图像进行增强处理,增强结果表明,本文改进的算法既能保留相干增强扩散的积极成果,又发挥了冲击滤波的锐化特性;然后,将本文算法应用到模拟的指纹识别算法中,选取合适的指纹图像作为模板指纹,实验证明,与相关算法相比,本文算法可以提高匹配度,在保护指纹图像细节点方面性能较好;最后,进行艺术创作,选取流状结构明显的图像作为实验对象,将本文算法、相干增强扩散和冲击滤波这三种方法的增强结果进行对比,并分析其在艺术创作等方面的特征表现。更多还原

【Abstract】 Automatic fingerprint recognition technology is now one of the most commonly used biometric identification technologies.The performance of the fingerprint recognition system is largely dependent on the quality of the input fingerprint.Normally,due to the influence of external environment and other factors,the noise generated during the fingerprint image acquisition will damage the fingerprint information and reduce the image contrast.Therefore,the degraded fingerprint image needs to be enhanced to facilitate the subsequent steps.This paper studies the fingerprint image enhancement method based on anisotropic diffusion,the main work is as follows:1.Sort out the relevant theories of fingerprint enhancement,and explain the importance of fingerprint image enhancement;2.Introduce the linear diffusion filtering process based on partial differential equations and the models of PM,Catte and FABD in nonlinear diffusion,and based on each anisotropic diffusion of edge enhancement and coherence enhancing diffusion algorithms,and at the same time,the shock filter algorithm is discussed;3.For the fingerprint image enhanced by coherence enhancing diffusion filter,there will be edge blurring and reduced contrast,and a new enhancement model is proposed : CED-SF model.The model uses the coherence enhancing diffusion algorithm to repair the degraded fingerprint image in the first stage,and then sharpens the edges by shock filter;4.To verify the effectiveness of the method in this paper,use MATLAB to perform simulation experiments and summarize the method in image enhancement And other three aspects.First,the three methods of this algorithm,coherence enhancing diffusion and shock filter are used to enhance the fingerprint image.The enhancement results show that the improved algorithm of this paper can not only retain the positive results of coherence enhancing diffusion,but also exert the sharpening characteristics of shock filter;Then,the algorithm of this paper is applied to the simulated fingerprint recognition algorithm,and the appropriate fingerprint image is selected as the template fingerprint.Experiments show that compared with the related algorithm,the algorithm of this paper can improve the matching degree and perform better in protecting the details of the fingerprint image;Finally,for artistic creation,select the image with obviousflow structure as the experimental object,compare the enhancement results of the three methods in this paper,coherence enhancing diffusion and shock filter,and analyze its characteristics in artistic creation and other aspects.更多还原