摘要
提出了一种基于边缘检测、数学形态学和颜色特征相结合的复杂背景下的快速车牌定位算法。首先,利用车牌区域具有较丰富的垂直边缘信息得到包含车牌区域在内的若干候选区域,然后利用车牌自身的三个结构特征从若干候选区域中提取车牌区域。仿真实验结果表明,该算法的准确率达到了91.2%。
This paper presents a fast license plate location algorithm under the complex background,this algorithm based on edge detection,mathematical morphology and color features.First,this paper get some candidate regions which including the license plate region through using license plate region has rich vertical edge information,then,three structure features of license plate are used to extract the license plate region from the candidate regions.Simulation experimental results show that this algorithm accuracy reached 91.2%.
出处
《工业控制计算机》
2013年第4期19-21,共3页
Industrial Control Computer
关键词
复杂背景
候选区域
车牌定位
complex background
candidate region
license plate location