指纹识别是生物特征识别技术中的一种，对于身份识别的准确性和可靠性的提高

指纹识别是生物特征识别技术中的一种，对于身份识别的准确性和可靠性的提高具有重要意义。指纹的唯一性和永久性使它运用在信息、公共、金融、社会管理等多方面。本文研究的就是指纹模式识别系统的算法及实现，其中包括指纹图像采集、图像预处理、特征提取和特征匹配等。随着指纹识别速度与准确性的提升，指纹识别将会在社会中得到越来越广泛的应用。
本文概述了指纹模式识别的预处理、指纹图像场及其计算。通过对指纹图像的分割、均衡、收敛、平滑、增强，以及指纹图像骨架的提取，实现纹图像特征提取与比对并构建指纹模式识别系统。指纹模式识别的预处理部分包含指纹图像畸变矫正、指纹图像场及其计算、指纹图像分割、指纹图像均衡、指纹图像收敛、指纹图像平滑、指纹图像增强、指纹图像二值化和指纹图像细化。
对于指纹图像畸变矫正，采取先构建自然模型，再构建物理模型，最后根据抽象的物理模型进行数学建模来对图像进行畸变矫正。对于指纹图像场及其计算，通过分别计算指纹图像强度场、指纹图像梯度场、指纹图像方向场最终计算出指纹图像场。对于指纹图像分割，通过将指纹图像灰度差的取值分布用灰度直方图进行标示来对指纹图像进行分割。对于指纹图像均衡，采取灰度均衡转换来进行指纹图像均衡。对于指纹图像的收敛，采用二维高斯函数描述指纹图像发散点的收敛函数的方法来进行收敛。对于指纹图像的平滑，通过指纹图像和模板算子的卷积来实现指纹图像的平滑。对于指纹图像的增强，本文采用Gabor滤波的方法对指纹图像进行增强处理。对于指纹图像的指纹图像骨架的提取，分为指纹图像二值化和指纹图像细化两个部分；对于指纹图像二值化，采取的方法是智能二值化领域分析法；对于指纹图像细化，本文使用的算法是查表法

結果 (英文) 1:

Fingerprint biometric recognition technology kind of identification is of great importance for the improvement of accuracy and reliability. The uniqueness of fingerprints and permanent so that it uses in the information, public, financial, management and other aspects of society. This article is the study of the algorithm and realization of fingerprint pattern recognition systems, including fingerprint image acquisition, image processing, feature extraction and matching, and so on. As the promotion of fingerprint recognition speed and accuracy, fingerprint recognition would have been widely used in our society.This article provides an overview of the preprocessing of fingerprint pattern recognition, image and its calculation. Through segmentation of fingerprint image, balance, convergence, smoothing, enhancement, as well as the skeleton extraction of fingerprint image, implementation patterns image feature extraction and comparison of construction and fingerprint recognition systems. Pre-processing of the fingerprint pattern recognition include fingerprints image distortion correction, image and its calculation, image segmentation, image balance, convergence, fingerprint the fingerprint image smoothing and fingerprint image binarization fingerprint image enhancement, and fingerprint image thinning.For fingerprint image distortion correction, first building a natural model, build physical models, based on abstract mathematical models of the physical model of image distortion correction. For fingerprint image field and calculate, by calculation fingerprint image intensity field, respectively the fingerprint image gradient, fingerprint direction field fingerprint images calculated field. Fingerprint image segmentation, by value of fingerprint image gray level difference distribution histogram for illustrations on fingerprint image segmentation. For fingerprint image balance, gray balance equilibrium transformations to fingerprint image. Convergence for fingerprint image using Gaussian function describes the divergence of the convergence of the fingerprint image function is used for the convergence. Fingerprint image smoothing through fingerprint image and template operator of convolution fingerprint image smoothing. Fingerprint image enhancement, using Gabor filtering method to fingerprint image enhancement processing. Fingerprint image skeleton extraction for fingerprint image into fingerprint image binarization fingerprint image thinning two parts; for fingerprint image binarization, takes the method of intelligent binary field analysis; for fingerprint image thinning, this algorithm is used by table look-up method