Image Defogging Algorithm for Smoke Conditions

Aiming at the problems of colour distortion and low saturation that occur in pictures taken under smoke conditions, this essay proposes an photo enhancement algorithm for smoke conditions, which improves the visual quality and readability of the images by combining the dark-channel a priori de-fogging method and the Retinex algorithm. The processed images are also analysed and evaluated by subjective and information entropy, mean gradient, mean square error and peak signal-to-noise ratio to verify the validity of the algorithm in this essay.In this essay, the dark channel priori method and Retinex algorithm are combined and improved, which has the following advantages over traditional photo enhancement algorithms:1.The colour and details of the image are restored more real: the dark channel a priori dehaze algorithm is able to restore the real colour and details of the image while removing the haze, reducing the colour distortion and detail blurring problems caused by the haze. Combined with Retinex algorithm can further improve the colour quality and detail clarity of the picture, making the photo more realistic and clear.2.More significant image enhancement effect: The image enhancement method that combines the dark channel priori defog algorithm and the Retinex algorithm can obtain more significant enhancement effects than when one algorithm is used alone. This method of using two algorithms in tandem improves the visual quality of the image by taking into account the need for fog removal and colour detail enhancement.