摘要
在HSV空间中,将车牌底色用1表示,非车牌底色用0表示,实现二值化。用二分法扫描滤波后的区域,对黑白变化特征进行识别,完成车牌粗定位。在灰度图像中对提取后的车牌区域进行直方图统计。大于或等于车牌字符灰度值的像素赋值为1,否则为0。然后进行特征提取及投影分析,实现车牌精定位。
In HSV space, plate background is represented by 1, others are represented by 0, to realize the binary. Complete plate positioned by using dichotomy scanning region after filtering, and identifying the characteristics changes of black and white. In the gray image, histogram statistics after extraction of license plate region. Gray value is greater than or equal to the license plate characters of pixel value is represented by 1, otherwise 0.Then the feature extraction and projection analysis, to realize license plate positioning.
出处
《电工技术学报》
EI
CSCD
北大核心
2015年第S1期527-530,共4页
Transactions of China Electrotechnical Society
基金
北京市优秀人才项目(2013D005017000017
2012D005017000004)
北京建筑大学博士启动基金(00331613002
0033161101)
国家自然科学基金项目(61402032)
北京市自然科学基金项目(4144072)
高动态导航技术北京市重点实验室(HDN2015102)