【摘要】 为消除明暗不均等低频干扰的影响 ,将字符成功地从车牌图像中提取出来 ,提出了基于小波分解的车牌图像分割方法 ,根据干扰信号与字符信号属于不同频域的特点 ,利用小波时频局部化分析的能力 ,在保留字符信息的同时滤除或抑制低频干扰成分 .其方法是先根据统计结果分析出字符和低频干扰信号所处的频率范围 ,然后对图像进行小波分解 ,过滤掉或抑制干扰信号所属的频段 ,再利用传统的聚类分析进行分割 ,这样就能够很好地消除白斑黑斑、光照不均匀、非车牌区域干扰等问题 ,即使图像灰度值反差很大也能将车牌字符很好地提取出来 .更多还原

【Abstract】 In order to get rid of the low frequency distribution like light and shade and obtain characters from license plate images, a way was put forward to segment license plate images based on wavelet decomposes. Because of the diffident frequency of license plate characters and disturbing signal, low frequency disturbing signal can be eliminated or restrained by wavelet transformation. Frequency part of character in license plate and the low frequency disturbing signal was found out according to the statistical result. Low frequency part was cleared up and the image segmented by traditional cluster analysis. The harmful effect can be eliminated and the letters segmented from the contrasty image of licenses plate.更多还原