基于FPGA的车牌识别系统的设计

Design of License Plate Recognition System Based on FPGA

车牌识别系统是一种能够检测到受监控路面的车辆并自动提取车辆牌照信息(含汉字字符、英文字母、阿拉伯数字及号牌颜色)进行处理的电子系统,其主要由图像捕获、图像处理和图像识别三部分构成。本设计基于ZYNQ平台实现车牌识别功能。其中在图像处理中实现了对车牌图像的形态学处理,包括二值化、定位、字符分割、缩放等。图像识别目前往往采用特征线识别和神经网络实现,而本设计中采用神经网络实现对字符图像的识别。通过本设计可以得出基于FPGA的小型化车牌识别系统对比传统车牌识别系统具有低成本、高性价比等特点的结论。

License plate recognition system is an electronic system that can detect vehicles on the monitored road and automatically extract vehicle license plate information(including Chinese characters,English letters,Arabic numerals and license plate color)for processing.It is mainly composed of image capture,image processing and image recognition.This design realizes the license plate recognition function based on ZYNQ platform.The morphological processing of license plate image is realized in image processing,including binarization,positioning,character segmentation,scaling and so on.At present,image recognition is often realized by feature line recognition and neural network,while neural network is used to realize character image recognition in this design.Through this design,it is concluded that the miniaturized license plate recognition system based on FPGA has the characteristics of lower cost and better performance compared with the traditional license plate recognition system.