A Novel Image Dehazing Method Using Adaptive Dark Channel Prior

To address the issue of suboptimal performance in different scenes by traditional dark channel dehazing algorithms, a novel image dehazing method based on an adaptive dark channel prior model Compared with traditional dark channel dehazing algorithms, this method is not affected by the environment and can achieve adaptive dehazing for various images, solving the problem of uneven final dehazing image quality caused by different image environments in traditional dark channel algorithms Firstly, the source image is calculated by using the dark channel prior, and an objective function for image dehazing is constructed utilizing the metrics PSNR and SSIM. Based on this, an improved particle swarm optimization algorithm is used to optimize the adaptive dark channel prior model for obtaining the threshold t0 of optimal transmission rate and the coefficients a and b of the objective function. Then, the optimal adaptive dark channel prior model is constructed to achieve the dehazing effect of different scene images. Finally, experimental results of six different scenes are presented to verify the feasibility and effectiveness of the proposed image dehazing method.