摘要
牌照读取过程大致可以分为预处理、牌照区域提取、牌照区域的二值化、字符分割、字符识别、后处理等几步。对于牌照区域的提取,综合利用了牌照的颜色信息和形状信息,取得了良好的提取效果;对于牌照区域的二值化,首先进行了通道的选择,然后计算阈值,最后二值化;在字符分割模块,分别在垂直、水平、垂直方向进行了字符分割,并计算了牌照的倾斜角,对倾斜程度比较严重的牌照,还要进行旋转(纠斜)处理;在字符识别模块,用最小欧氏距离法进行识别,对‘8’和‘B’及‘0’和‘D’等易混淆的字符进行特殊处理;在最后的后处理模块中,进行了包括机器检查和人工检查在内的合法性检查。实验结果显示,基于彩色图像处理技术的车辆牌照自动读取系统具有较高的识别速度和精度。
A car plate reading system can be divided into the following steps： preprocessing, plate region extraction, plate region thresholding, character segmentation, character recognition and post-processing. For plate region extraction, a combination of the colour and shape information of plate is used and a satisfactory extraction result is achieved. For plate region thresholding, first a channel is selected, then the threshold is computed and finally the region is thresholded. For character segmentation, the character is segmented along vertical, horizontal and vertical directions and the sloping angle is also computed. For a seriously sloped plate, rotation（anti-slope） is needed. For character recognition, the minimum Euclidian distance is used. And for those confusing characters such as ‘ 8’＆ ‘ B’ and‘ 0 ’＆‘D’ , a special processing is necessary. And finally, for post-processing, validity is checked by machine and manually. The experimental results indicate that our automatic car plate reading system based on colour image processing is quite quick and accurate.
出处
《计算机工程与科学》
CSCD
2005年第9期47-50,共4页
Computer Engineering & Science
