基于混合远景和近景区域暗通道算法的图像去雾研究

Image Dehazing Based on Near and Far Scene Dark Channel Prior

为了提高图像去雾效果,提出混合远景和近景区域暗通道算法。首先建立远景和近景区域暗通道,为近景区域最小值通道和远景区域最小值通道之和,不同区域设置不同的调节系数;接着估计大气光值,通过混合暗通道远景区域的前50个像素点颜色值的平均值获得,增设校正系数;然后导向滤波,采用自适应滑动窗口优化透射率;最后给出了算法流程。实验仿真显示所提算法去雾图像视觉效果较好,定量分析指标对比度、可见边对比度值较优。

In order to improve the effect of image dehazing,near and far scene dark channel prior,was proposed.Firstly,mixed dark channel prior was established,which is included minimum channel of near scene and dark channel prior of far scene,and set adjustment coefficient of different scene region.Secondly,atmosphere light value was estimated with 50 pixel color average value of mixed dark channel prior far scene,and added correction coefficient.Thirdly,transmission was optimizated with adaptive sliding window of guided filtering.Finally,process was given.Experiment simulation showed that mixed dark channel prior has better visual effect,and contrast ratio and visible edge contrast ratio of quantitative analysis indexes are better than other algorithms.