基于色彩分割与体态纹理分析的车牌定位方法

介绍了一种基于色彩分割、体态分析及数学形态学纹理分析的车牌定位方法。该方法利用车牌与背景的色彩特征．在HSV空间内根据车牌颜色的三分量取值范围和色彩距离阚值进行色彩分割以初步过滤背景。对分割后得到的与车牌底色相符合的连通区域再根据车牌体态特征和车牌字符纹理特征，应用数学形态学处理及连通域体态分析等手段，逐步过滤干扰区域，最终定位正确的车牌位置。该方法充分考虑了车牌与背景的色彩、体态及纹理特征的差别，实验证明在复杂背景下具有很强的适应性与鲁棒性。

A Method for License Plate Location Based on Color Segmentation and Shape-texture Analysis

This paper introduces a new license plate location method,which based on color segmentation,license plate shape analysis and mathematic morphology analysis.In HSV color space,the method employs color segmentation algorithm,which utilizes three elements' value range of the license plate color and a color distance threshold,to the image to filtrate background prepartorily.And then according to the shape and texture features,mathematics morph operation and continuous areas' shape analysis are applied to the continuous areas that fulfill the color property of the license plate to gradually filtrate disturbed areas.The method fully considers the difference of the color,shape and texture features between license and background,and has been proved to be robust and highly adaptive in complex environment in actual application.