计算机科学 ›› 2009, Vol. 36 ›› Issue (8): 296-299.
• 图形图像及体系结构 • 上一篇 下一篇
片兆宇,孟祥萍,张红
出版日期:
发布日期:
PIAN Zhao-yu MENG Xiang-ping, ZHANG Hong
Online:
Published:
摘要: 车牌定位是车牌识别系统中的关键环节。针对车牌图像区域丰富的边缘信息以及独特的特征,提出一种实用而有效的车牌定位方法。首先,将原始图像转换到HSI空间上,利用图像的字符边缘特征信息和数学形态学操作对目标区域进行粗定位;然后,根据车牌固有的特征定义4种不同的特征值,通过自组织神经网络的训练,实现对车牌区域的精确定位。最后,对80幅彩色图像进行测试,其有效率可达96. 25%.
关键词: 车牌定位,HSI空间,形态学,自组织神经网络
Abstract: The location of license plate plays a crucial role in license plate recognition system. Based on the rich edge information and unique features of the license plate region,a practical and effective method of locating the plate was proposed. Firstly, the original image was transferred to HSI space, and the rough locations were detected by using the edge information of the characters and the operation of the mathematical morphology; and then, the accurate region of license plate was decided by the SOM using four features proposed by this paper. Finally, the experiments on 80 color images show that the efficiency of the proposed algorithm is 96. 25%.
Key words: License plate location, HSI space, Morphology, Self organizing map
片兆宇,孟祥萍,张红. 多阶段车牌定位算法[J]. 计算机科学, 2009, 36(8): 296-299. https://doi.org/
PIAN Zhao-yu MENG Xiang-ping, ZHANG Hong. License Plate Location Algorithm Based on Multi-phases[J]. Computer Science, 2009, 36(8): 296-299. https://doi.org/
0 / / 推荐
导出引用管理器 EndNote|Reference Manager|ProCite|BibTeX|RefWorks
链接本文: https://www.jsjkx.com/CN/
https://www.jsjkx.com/CN/Y2009/V36/I8/296
Cited