车牌识别中字符分割算法的研究与实现的毕业论文2013.docVIP

毕 业 论 文 题 目 摘 要 车牌自动识别系统是指通过计算机视觉、图像处理与模式识别等方法从车辆图像中提取车牌字符信息的系统，从而确定车辆身份，是实现交通管理智能化的非常重要的环节之一。研究车牌自动识别技术是具有重要实用价值的。 车牌自动识别系统一般包括车牌定位、字符分割和字符识别三部分。本文主要针对车牌字符分割进行研究，研究了与其相关的分割算法，进而实现车牌字符的分割。 首先，对车牌字符分割的预处理进行了较深入研究，尤其是车牌图像的边缘检测[2]。考虑到车牌图像的特征，给出了图像灰度校正，平滑处理以及边缘提取这一系列方法对图像进行了预处理。其次，在车牌定位中，根据车牌中的字母与数字具有连通性的特点，给出一种基于数学形态学的车牌区域定位方法，并对车牌字符进行了准确定位。实验结果表明，该方法可以有效地对车牌进行定位。然后，考虑到车牌本身特征，以及车牌中图像像素分布的特点，给出了一种基于车牌像素和模板匹配相结合的字符分割方法。通过实验表明，这种方法的字符分割结果比较准确。 关键词：车牌定位；预处理；字符分割 ABSTRACT Automatic license plate recognition(LPR) is the technique to draw the character information from the vehicle image and confirm the identity of the vehicle through using computer vision，image manipulation,pattern recognition and on．LPR is a important study field in the ITS，and it has great apply value． Automatic license plate recognition system generally consists of license plate location, character segmentation and character recognition three parts. In this paper, license plate character segmentation research, and put forward the related algorithm, and then realize the license plate character segmentation system. Firstly, the pretreatment of the license plate character segmentation is a more in-depth research, especially the license plate image edge detection. Taking into account the characteristics of the image plate, the image is given gradation correction, smoothing processing and edge extraction method in this series of preprocessing the image. Secondly, in license plate location, in accordance with the license plate letters and numbers with connectivity features, presents a mathematical morphology based license plate location method, and the license plate characters were accurate positioning. Experimental results show that this method can effectively locate the license plate. Then, taking into account the characteristics of the license plate itself, as well as the license plate of the image pixel distribution characteristics of a given pixel based on license plate character segmentation m