基于机器视觉的车牌识别系统设计

Design of License Plate Recognition System Based on Machine Vision

车牌识别系统是智能化交通管理系统的核心部分。介绍基于机器视觉的车牌识别系统设计的过程,车牌识别步骤为:车牌图像采集、图像预处理、车牌定位与提取、车牌字符的分割和字符识别。由于实际生活中视觉传感器所采集的车牌图像可能模糊不清,图像容易受到光照条件、阴雨天气等影响,导致车牌图像本身含有噪声干扰,需要运用数字图像处理算法减小噪声干扰,便于正确识别车牌字符。该设计包涵采用MATLAB软件仿真测试、GUI人机界面构建、通过图像导入和各个环节的图像处理,最终实现车牌号码的识别。实验结果验证所用方法的可行性。

License plate recognition(LPR)system is the core part of intelligent traffic management system.The design process of license plate recognition system based on machine vision was introduced.The steps of license plate recognition are:license plate image acquisition,image preprocessing,license plate location and extraction,license plate character segmentation and character recognition.In real life,the license plate image captured by the vision sensor may be blurred,and the image is easily affected by illumination conditions,rainy weather and so on.As a result,the license plate image itself contains noise interference,it is necessary to use digital image processing algorithm to reduce noise interference and facilitate the correct recognition of license plate characters.The design includes the use of MATLAB software simulation test,GUI man-machine interface construction,through image import and image processing of each link,and ultimately realize the license plate number recognition.The experimental results verify the feasibility of the method.