摘要
车牌定位是车辆牌照识别的重要基础。传统的车牌定位方法由于受到光照变化、图像尺寸大小或者车身颜色与车牌颜色相近的影响具有局限性。该文提出了一种基于纹理分析和字符特征相结合的方法，重点介绍了图像预处理、梯度化、纹理分析和字符特征分析等主要步骤并给出了实验结果。
License plate location is the base of license plate recognition. Traditional methods have limitations in light change,image size dif- ferences and those plates which have the similar color with the car. The paper proposed a method based on texture and character features. It introduces image processing,image gradient,texture and character features analysis with experimental results.
作者
魏中衡
张帅
WEI Zhong-heng, ZHANG Shuai （Shandong University of Science and Technology, Qingdao 266590, China）
出处
《电脑知识与技术》
2012年第8期5434-5437,共4页
Computer Knowledge and Technology
关键词
车牌定位
纹理分析
字符特征
sobel变换
license plate location
texture analysis
character features
sobel transform
