本文提出了一种基于门限化和宽线检测算子的自适应牌照定位方法 ,解决了复杂背景下的汽车牌照定位问题。利用汽车头部图像的一些知识信息 ,采用自适应门限方法 ,辅以宽线检测算子进行后验校正 ,能够准确地分割出汽车牌照 ,然后对牌照中的汉字和数字进行分离 ,正确率达 95 %以上。 In this paper, an adaptive license plate location method based on threshold and wide line detection operators is proposed to solve the problem of vehicle license plate location under complicated background. Using some knowledge of car head image, using adaptive threshold method, supplemented by a wide line detection operator for posterior correction can accurately segment the license plate, and then the license plate in the separation of Chinese characters and numbers, the correct rate of up to above 95.