基于加权平均值的汽车车牌监控图像过滤技术研究

车牌监控图像由于照明、天气、运动目标位置和运动目标速度的不同,图像质量差异很大,从而不利于车牌监控的定位和识别。本文采用最大值法、平均值法和加权平均值法三种方法对车牌监控图像进行过滤,结果表明:采用加权平均值法进行车牌图像颜色过滤能够保留绝大部分的汽车车牌信息,使得目标和背景之间边界清晰,是一种较好的车牌图像彩色过滤方法。

video surveillance images due to different fighting, weather, target position and movement speed of the moving object, a large difference in image quality, there are ot^en low in contrast, gray uneven distribution, noise, etc., to the detriment of video surveillance and positioning identified. In order to eliminate these negative factors , we use the image color filtering technology, video surveillance images for color filters . This paper describes the principles of necessity and video color filters, and then to monitor the license plate image, for example, describes the process of video surveillance images, and uses three color filter method maximum method, the mean and weighted mean method its filter results were verified. The results show that using the weighted average method method method for video image color filter to retain the majority of car license plate information, and background information on unrelated effectively clear filter between the target and the background border conducive license plate recognition is a species better image color filter method.