中文摘要：

针对去雾方法存在运算复杂度高、图像边沿细节模糊等缺点，提出一种基于K值邻域均值滤波的去雾算法。利用K值邻域均值滤波算法对大气透射图进行估计，通过大气散射物理模型进行去雾处理，最后通过乘法运算实现图像的亮度调整。实验结果表明，K值邻域均值滤波算法具有更快的处理速度，满足视频图像去雾处理系统的实时性要求，且解决了图像去雾后边沿模糊的问题，去雾效果更好。

英文摘要：

In view of high computational complexity, fuzzy image edge details and other shortcomings of the defogging meth od, a defogging algorithm based on K value domain average filtering method is proposed. The atmospheric transmission is estimated by using K value domain average filtering method. Haze is removed by the physical model of atmospheric scatter- ing. Finally the brightness enhancement of image is realized by multiplicative operation. Experimental results show that the K value domain average filtering method has faster processing speed to meet the requirements of real time video image de fogging system. It resolves the problem of image edge blurring after defogging and gets better defogging effect.