Regularization model based on transmission constraint

Abstract

Aiming at the problems of blocking and artifacts in traditional dark channel prior dehazing algorithms, an image dehazing algorithm based on dark channel priors is proposed. First, the transmission is corrected once according to the characteristics of the dark channel of the image to solve the block effect caused by the underestimation of the transmission. Secondly, in order to overcome the problem of blurring of details, this paper proposes a regularized model based on transmission correction to optimize transmission twice to overcome the artifact problem in traditional dehazing algorithms. In addition, the Alternating Direction Method of Multipliers (ADMM) is used to solve the new model. Finally, the atmospheric scattering model is used to restore the image. Numerical experimental results show that the proposed method is significantly better than the traditional image dehazing algorithm. Taking the $480\times 540$ Boat image as an example, the PSNR and SSIM value of the method in the article are improved by 1. 8dB and 0. 56dB on average. The algorithm proposed in the article solves the blocking effect while eliminating the haze, and eliminates the artifacts of the image. It is significantly better than the traditional dehazing algorithm in terms of visual effects and objective evaluation indicators.