本文的研究背景是车牌字符识别，识别包括0-9和A-E及R，共16个字符。每个灰度图像都是由36×20个像素点组成，存储形式为36×20的矩阵。将灰度图二值化后，形成0、1矩阵。通过一些简单的图像预处理后，采用一种新的面积法进行特征提取。-This paper is the research background of license plate character recognition, identification including 0-9 and A-E and R, 16 characters. Each gray image is the 36 × 20 pixels composition, storage form of 36 × 20 matrix. Will gray image after binary, forming a 0, 1 matrix. With some simple image preprocessing, adopting a new method of feature extraction.