Image fusion dehazing algorithm based on multi-logarithmic transform

In hazy scenarios, suspended particles in the atmosphere will absorb and scatter the transmitted natural light, resulting in a serious degradation of image quality obtained by imaging equipment, which greatly affects the visual perception of images. Aiming at the problems such as low contrast, color distortion and lack of detail information in areas with high haze concentration in images acquired by image acquisition equipment in hazy days, this paper proposes a dehazing algorithm based on exposure image fusion based on multi-logarithmic transform, which improves image quality while effectively dehazing images and avoids the edge effect in the sky part of images after dehazing. Firstly, the original hazy image was transformed by logarithmic multiple times to produce multiple images with different exposure to be fused. Then, all the input images with logarithmic transformation and weight graphs were fused by multi-scale pyramid fusion method to obtain the dehazing image. In order to verify the effectiveness of the algorithm in this paper, the results of the proposed algorithm and six mainstream image dehazing algorithms are compared in two aspects: subjective evaluation and objective evaluation. Experimental results show that the image processed by the proposed algorithm presents better visual effects than that processed by other algorithms. The proposed algorithm can effectively improve image contrast, improve image distortion, improve the visibility of detail information in areas with high hazy concentration, and the scenery color is natural. Good results are obtained under two objective evaluation indexes of image quality, namely peak signal-to-noise ratio and structural similarity, which further proves that the algorithm proposed in this paper has good dehazing performance, can effectively improve the image visibility, and has a good overall color preservation degree of the image.