说明：  汽车车牌的图像识别。 根据车牌凸角轮廓特征，采用MPP算法进行车牌倾斜校正；采用彩色图像分割方法，最大限度的去除图像噪声干扰；设计了基于神经网络的多种彩色背景车牌字符识别方法，取得了良好的效果。
(Image recognition of vehicle license plates. According to plate convex contour feature, the use of MPP plate tilt correction algorithm using color image segmentation method, the maximum removal of image noise designed based on a variety of colored background neural network method for license plate recognition, and achieved good results.)