摘要
本文根据车牌丰富的纹理特征,提出了一种准确、快速的汽车牌照定位方法。该方法首先利用车牌的纹理特征,采用模板匹配的方法进行初步定位;然后,根据细纹理特征和牌照的先验知识,综合利用数学形态学、投影以及颜色信息判别的方法进行精确定位。同时,针对实际应用的要求讨论了对其中的一些方法的改进。大量的实验结果表明,该方法具有快速、准确率高的优点。
This paper proposes a method for locating car license plates using the abundant texture features. Firstly, we obtain a number of candidates based on the texture characteristics of license plates using template matching. Then the car license location is distilled from the candidate areas simultaneity using the mathematical morphology, projection and color information according to the texture and foreknowledge feature of license plates. Some improvement methods are also discussed in this paper so as to satisfy the application. The experimental results show that the proposed method is quite effective and rapid.
出处
《计算机工程与科学》
CSCD
2008年第8期39-41,82,共4页
Computer Engineering & Science
基金
西安市科委科技成果转换项目
关键词
车牌定位
模板匹配
纹理分析
数学形态学
license plate localization
template matching
texture analysis
mathematical morphology
