A Quality Evaluation Method for Images in Adverse Weather Conditions

By adopting an adaptive image quality evaluation method under adverse weather conditions, an advanced image defogging system has been developed. This system aims to predict the optimal algorithm for achieving the best image defogging results. The system integrates nine efficient defogging algorithms. To comprehensively evaluate the defogging performance of these algorithms, a set of evaluation metrics including PSNR, SSIM, LPIPS, ENL, MSE, and RMSE is employed to establish a comprehensive evaluation method. By calculating the values of these evaluation metrics, the optimal algorithm suitable for the image is predicted, and the defogged image produced by the optimal algorithm is processed to output the best-defogged image finally. All predicted data is stored in a database. When defogging an image, if there is matching data in the database, the corresponding optimal algorithm and image processing are directly selected, thereby enhancing the overall defogging efficiency.