Study on Outdoor Fog Detection and Image Dehazing Algorithm

This paper studies outdoor fog detection and image dehazing algorithms. Firstly, a random forest method is used to classify images into foggy or non-foggy based on the dark channel gray value distribution and saturation features of images. Then, an improved AOD-Net algorithm is used to remove haze from the foggy images. Finally, the effectiveness of the proposed methods is verified by experiments. Experimental results show that the random forest classifier based on the dark channel and saturation features can accurately detect foggy images, while the improved AOD-Net algorithm can effectively remove the haze in the images and improve their clarity and visibility. The proposed methods have practical value and can be widely used in aerial photography, security monitoring, and other fields.