| 一种基于数学形态学的车牌图像边缘检测及提取算法 |
| |
| 引用本文： | 郭亚,宋焕生.一种基于数学形态学的车牌图像边缘检测及提取算法[J].交通与计算机,2009,27(5):158-160. |
| |
| 作者姓名： | 郭亚 宋焕生 |
| |
| 作者单位： | 1. 江西理工大学机电工程学院,江西,赣州,341000
2. 长安大学信息工程学院,西安,710064 |
| |
| 摘 要： | 针对牌照字符目标与背景的差异性，提出1种基于数学形态学的牌照字符提取方法。该算法对输入图像进行相邻像素差分，保留纵向纹理；通过二值化和数学形态学处理，得到许多大小不同的连通域；计算各连通域的特征参数，利用牌照区域和背景区域对应的连通量的特征差别，实现有效抑制背景而保留牌照的目的。选择了1组有代表性的图像作为测试样本，验证了该算法的准确性和环境适应性。
|
| 关 键 词： | 交通工程 数学形态学 二值化 连通域 边缘检测 |
A New Algorithm for License Plate Edge Detection and Extraction Based on Mathematical Morphology |
| |
| GUO Ya,SONG Huansheng.A New Algorithm for License Plate Edge Detection and Extraction Based on Mathematical Morphology[J].Computer and Communications,2009,27(5):158-160. |
| |
| Authors: | GUO Ya SONG Huansheng |
| |
| Institution: | Jiangxi University of Science and Technology;Ganzhou 341000;Jiangxi;China;Information Engineering College;Chang'an University;Xi'an 710064;China |
| |
| Abstract: | According to the differences between license plate characters and their background,an algorithm based on mathematical morphology is proposed to extract license plate characters.First,the used image is processed by difference method between adjacent pixels in the inter-rows,thus vertical textural information is preserved.Then,the image is processed by binaryzation and morphology to obtain many connected regions.These connected regions are then labeled and their feature parameters are calculated.As the connec... |
| |
| Keywords: | traffic engineering mathematical morphology binarization connected regions edge detection |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
