我国是较早意识到指纹重要性的国家，早在宋朝就有了“画押”一说。虽然指纹识别技术在我国起步的很早，但是由于当时条件的限制，使得这一技术并没有得到有效的推广。如今，科技变化日新月异，带动了这一技术的发展。 指纹识别技术主要包括：指纹图像预处理、特征点提取和特征点匹配等过程，其中指纹图像预处理过程包括图像的平滑处理、计算图像的方向图、图像的分割处理、图像的增强处理、图像的二值化处理和图像的形态学等处理等[1]。对指纹图像进行平滑处理的目是为了消除指纹图像在传输或者采集过程中出现的模糊问题，确保指纹图像的细节特征提取的准确性和可靠性，从而提高了指纹识别的正确性。指纹图像细化的目的是为了使提取的特征点的位置更加精准，以便给指纹匹配打好基础。 虽然指纹识别技术发展已经趋于成熟，但是仍然存在许多有待于解决、有待于优化的问题。这些问题普遍存在于指纹采集的过程、指纹预处理过程、指纹特征点提取过程、指纹匹配的过程之中。本论文在对指纹识别主流算法进行阐述的基础上，分别针对指纹预处理、指纹分割技术、指纹匹配技术三个过程各提出一种优化算法，用于解决生活中遇到的问题。 本论文的算法分四步完成，分别在第三到第六章中有所体现，每一章都有一种完整的算法，代表着指纹识别算法的某一步骤，四章算法相结合即为指纹识别算法。 本论文所提出的算法均在MATLAB上进行了仿真，并且在附录中有相应的程序和截图，从而确保论文的真实性和算法的优越性。通过仿真结果可以发现，本文提出的算法在主观评价指标和客观评价指标上均符合要求。

The importance of fingerprint was earliest known by china. In Tang Dynasty, people was compelled to press their fingerprint in paper. Even we knew fingerprint for a long time, but we didn't expanded this technology widely due to some limits. However, modern since has prompt this technology fast applied into our life. General process of the fingerprint identification technology mainly include fingerprint pretreatment, feature extraction and fingerprint matching .Pretreatment of the fingerprint image include fingerprint image Smoothing, calculation direction figure, image segmentation, image enhancement, binary image and refined. The purpose of the fingerprint image enhancement is to avoid false characters information generated by the image, ensure the accuracy and reliability of the details of the feature extraction. The purpose of fingerprint image thinning is to extract image feature points and to finish the subsequent work. Now, the technology of fingerprint has developed into a relatively mature stage, but there still are many problems waiting to be solved and optimized. These problems prevalently exists in many phase of fingerprint recognition ,such as fingerprint acquisition ,fingerprint pre-processing, fingerprint feature point extraction and fingerprint matching. Based on the description of mean stream fingerprint recognitions, three optimizing methods are proposed in the procedure of finger point pre-processing, finger point division and finger point matching respectively in order to solve the common problems in daily life. The algorithm in this paper is complished by four steps which are reflected respectively from chapter three to chapter six. Every chapter has one complete algorithm, representing one step of finger point recognition algorithm, and the combination of these four algorithms will be complete algorithm. In this thesis, experiments have been done in the PC of software MATLAB and experiment results indicate that the fingerprint verification algorithm is more effective, feasible and it achieves the desired results.