车牌自动识别系统的嵌入式设计

Embedded Design of License Plate Recognition System

科技正以前所未有的速度改变着人们的生活,从而影响到了交通系统的发展。为了组建更加智能化、人性化的交通系统,车牌自动识别系统应运而生。本文采用了嵌入式系统作为车牌自动识别系统平台。通过对硬件和软件的探讨,搭建出车牌自动识别系统的框架。使用Linux作为操作系统平台,使用OpenCV开源函数库进行图像的分析与处理,在这些基础上完成了车牌识别系统的整体模型。首先对车牌进行灰度变换,然后选取阈值使图像二值化,接着进行车牌的定位,获得车牌图像后进行车牌字符的切割,使单个字符分开,最后对分割后的字符和标准的字符模板进行匹配,从而获得车牌信息记录下来并传输到需要的主机上。

The technology is changing people＇s life at unprecedented speed, also affecting the development of traffic system. In order to form a more intelligent and humanized traffic system, the license plate recognition system emerged. This paper used the embedded system as a license plate recognition system platform. Through discussing the hardware and software, this paper builds framework of the license plate recognition system. It uses Linux as operating system platform, and uses OpenCV open source function library to conduct image analysis and processing. On the basis, the whole model of the license plate recognition system is completed. First, the system conducts gray transform for vehicle license plate, and selects threshold to make the image binary, then locates the license plate, cuts and divides plate character after receiving a license images, and then matches them with standard characters formwork, and gets the license plate information and records and transmits to the needed hardware.