摘要
随着科技的发展,车牌识别系统得到了很多应用,文中采用投影法进行倾斜校正,用不同的方法对光照不同的图像分别进行二值化处理,字符识别部分采用模板匹配及特征点匹配相结合的方法,提高了识别速度和准确性。文章所采用的合理且实用的图象处理方案,能够使车牌识别系统获得理想清晰的车牌图像。
License plate recognition system makes a good application in the development of science& technology. Projection method is taken to search the lean of a picture, Otsu method is used in the binarization process aimed at images which have well-proportioned Illumination, and Bemsen method is used in the images which are not well-proportioned. A method which matches template matching with characteristic of dots well is applied in this article to process recognition of characters, the processing rate and accuracy is proved advanced a lot through this method. All the image processing methods we adopt in this article proved adaptable and can get a relatively high quality license image.
出处
《微计算机信息》
北大核心
2007年第28期241-243,共3页
Control & Automation
基金
国际自然科学基金项目(60641010)
关键词
车牌识别
图像处理
二值化
模板匹配
License Plate Recognition
image processing
image binary
template matching