To improve the real-time and accuracy of intelligent vehicle character recognition algorithm, this paper mainly studies a new method for license plate location and character cutting: license plate location is carried out by combining HSV color space recognition with Sobel edge detection. The outer rectangle of the outermost contour of the character is used to judge the character boundary and segment the character. First, the original picture is converted to the HSV space, then the color information of the license plate is combined with the edge information to locate the license plate roughly, and finally the specific position of the license plate is determined according to the aspect ratio of the license plate. The location of the license plate is extracted for character segmentation preprocessing. Then, by traversing, the contour which meets the aspect ratio condition is found and the corresponding circumscribed rectangle is taken. The coordinate information of the rectangle is extracted and stored as a matrix, which is the basis of character segmentation. It is proved that this method has better accuracy and reliability than the traditional methods.

为提高智能车辆字符识别算法的实时性和准确性，本文主要研究了一种车牌定位和字符切割的新方法：将HSV颜色空间识别与Sobel边缘检测相结合进行车牌定位。字符最外轮廓的外矩形用于判断字符边界和分割字符。首先将原始图片转换为HSV空间，然后结合车牌的颜色信息和边缘信息对车牌进行粗略定位，最后根据车牌的纵横比确定车牌的具体位置车牌。提取车牌位置进行字符分割预处理。然后，通过遍历，找到满足纵横比条件的轮廓，取对应的外接矩形。矩形的坐标信息被提取出来并存储为一个矩阵，这是字符分割的基础。实践证明，该方法比传统方法具有更好的准确性和可靠性。