摘要
车牌定位是车牌识别系统中最重要的步骤，针对车牌区域的特点，提出一种简单快速的车牌粗定位方法。首先对拍摄的包含车牌的彩色车身图像进行灰度化，并采用自动阈值方法进行二值化，对于不同颜色车身图像都能得到理想的二值图像，然后对二值图像利用梯度进行边缘检测，再根据车牌区域水平方向具有黑白灰度跳变较为频繁的特征，采用行扫描图像的灰度跳变频率来定位车牌区域。在拍摄的车牌图像上进行定位实验表明，方法具有简单实用，速度快的特点。
Abstract
Automatic license plate location is the most important step in license plate recognition system. According to the characteristics of the li-cense plate area, proposes a simple and fast method for the coarse location of the license plate. First of all, the color image of the license plate is gray, and the automatic threshold method is used to carry on binarization. The method can be used for different color images. Then gradient operator is used to detect the edge of the binary image. And then the gray-hopping frequency is used to locate the license plate area. Experiments on license plate images show that the proposed method is simple, practical, and fast.
NETL
NSTL
维普
万方数据