数字图像处理应用于车辆牌照识别的研究

Identification of License Plate by Digital Image Processing

探讨了将数字图像处理技术应用于机车车辆牌照识别的基本方法.在图像预处理部分,选用灰度差倒数加权平均算法对源图像进行滤波处理;应用分段线性变换算法增强图像对比度.通过边缘检测算法对车辆牌照进行准确地定位及对文字进行分割,完成对车牌上的字符的识别是利用各个字符的图像和字体的特征值,对其中每个字符进行有效、准确的识别.

Digital image processing was applied to the recognition system of license plate. In this system, the original image of license plate was firstly transformed, compressed, enhanced and horizontally rectified, then fixed in a position and cut apart by verge exam process. The characteristic extracting and model matching was used in character recognition. Verge exam arithmetic was used to correctly locate license plate and cut apart characters, and every character was recognized accurately and effectively through applying feature value of it.