汽车牌照自动识别中二值化方法的研究

在车牌图像处理过程中，首先遇到的问题就是图像二值化．针对汽车牌照的特点，介绍了一种新的二值化阈值方法，此方法借鉴了P片法的二值化思想，不仅充分考虑了车牌自身的特点，同时融入对直方图形态的考虑，能准确找到双峰直方图的波谷位置．并与Otsu算法和另一种全局动态阈值算法作了比较，从结果上看执行效果较好．

Research on binarization method of license plate recognition in automatic recognition

In the course of dealing with license plate image,the binarization is significant. Directed against the characteristic of the license plate, a new method of taking threshold value of binarization is introduced. The new method draws lessons from the thought of binarization in P-titl and fully considers the characteristic of the license plate and the shape of histogram, so the lowest spot in bimodal histogram is found accurately by using this method. Otsu method and another method of globally dynamic threshold are compared with this method.The results indicate that the new method is better.