论文中文题名: | 基于kitk1201的指纹识别系统研究 |
姓名: | |
学号: | 201106234 |
保密级别: | 公开 |
学科代码: | 081104 |
学科名称: | 模式识别与智能系统 |
学生类型: | 硕士 |
学位年度: | 2014 |
院系: | |
专业: | |
第一导师姓名: | |
论文外文题名: | Fingerprint Identification System Research Based on the Kitk 1201 |
论文中文关键词: | |
论文外文关键词: | Fingerprint Recognition ; preprocessing ; image enhancement ; binarization ; Feature |
论文中文摘要: |
指纹是当今应用很广泛的生物特征之一,指纹的独特性、稳定性和防欺骗性,使得指纹识别在身份识别中有很重要的作用,成为当前应用广泛的身份验证和识别技术之一。它涉及模式识别、数字图像处理等领域,广泛应用于信息安全,犯罪识别等众多领域,有着重要的理论意义和应用价值。
本文对指纹预处理过程进行系统研究,针对一般指纹预处理算法对指纹采集仪kitk1201采集的指纹图像处理时指纹图像中间部分信息会丢失的问题,提出一种改进的预处理方法。该方法分别采用灰度阈值法和高低帽变换法对指纹图像进行分割,将其结果图像叠加,避免了指纹特征信息的丢失,使得指纹纹线变得清晰、连续、光滑,改善了指纹图像的质量。在指纹匹配方面为了提高指纹识别的效率,提出一种基于特征点的二级指纹匹配算法。第一匹配阶段利用奇异点构造特征点的局部特征向量来进行的匹配,实现快速排除纹线差异明显的图片,以此来缩小匹配的范围;第二阶段以初匹配的结果为基础进行特征点集的坐标校准,对校准后的点集进行匹配,得到匹配结果并且进行判决。
设计了实验室人员管理系统。实验表明,本文提出的指纹预处理方法和指纹匹配方法,应用于kitk1201指纹采集仪构成的指纹系统,识别效果良好,降低了指纹匹配的误识率,缩短了匹配时间。
﹀
|
论文外文摘要: |
Fingerprints is one of the widely used biological characteristics with it’s uniqueness, stability and anti- deceptive, making fingerprint identification has a very important role in the identification, and become one of the broad used authentication and identification technology . It relates to the field of pattern recognition, digital image processing, etc., are widely used in many fields of information security, criminal identification, also has important theoretical significance and application value.
The whole process of fingerprint preprocessing was analyzed, for the problem of the fingerprint image information is lost in the middle part of fingerprint scanner kitk1201 with general fingerprint preprocessing algorithm, an improved pretreatment method is proposed. The grayscale threshold method and top-hat and bottom-hat converted method were used for fingerprint image segmentation in the method, then the result is superimposed ,which avoids the loss of fingerprint feature information, and makes fingerprint ridge clear, continuous, smooth. It can improve the quality of the fingerprint image. In fingerprint matching aspects, this paper proposes an based on feature points two fingerprints matching algorithm, in order to improve fingerprint recognition efficiency. The first phase of the match, which is matched by the use of local structural feature vectors of singularity points, the ability to quickly exclude those fingerprint ridge clearly inappropriate pictures, in order to narrow the scope of matches; The second stage , it results as the basis for beginning the matching set of the calibration of feature point coordinates, the calibrated points are matched, and the matching results obtained judgment result.
laboratory personnel management system was designed, Above experiments show that the proposed pretreatment method and fingerprint matching method have good recognized effect for fingerprint scanner kitk 1201 constitute fingerprint system, fingerprint matching reduces the error rate, shorten the matching time.
﹀
|
中图分类号: | TP391.41 |
开放日期: | 2014-06-18 |