中文摘要：

根据彩色图像RGB三色学原理及室间域的图像平滑算法中的Gauss低通滤波器对图像进行预处理，利用OTSU法来求取图像的19值，并在此基础对图像进行二值化分割，图像分割是车牌自动定位技术中的基本问题之一，也是车牌自动识别过程中的关键步骤。实验证明：只要合理选择其该方法中的参数，正确率可达到95％以上，且处理时间小于0．5s。

英文摘要：

According to the RGB three-color principle and the smooth algorithm in image processing, this paper uses Gauss low-pass filter in spatial domain to the pre-processing of image, and adopts OTSU to get threshold of the image. And then, based on OTSU, the image is segmented into binary one, which is one of the basic problems of the license ptate auto-registration technology, which is also the key step of the license plate auto-recognition process. Experiment shows that the rate of correctness of 95 above and the processing time of less than 0.5s can be obtained so far as the reasonable parameters are selected in this method.