A Novel Image Quality Assessment Method for Dehazed Image

Abstract

In order to measure the effectiveness of dehazing algorithm, we need an objective and fair defogging quality evaluation algorithm to evaluate the effect of image defogging. Image quality is an important metric to compare the performance of defogging algorithms and optimize system parameters. At present, Image Quality Assessment (IQA) methods can be divided into subjective evaluation method and objective evaluation method. In view of the extensive application of Structure Similarity Index Measure (SSIM) and Peak-Signal to Noise Ratio (PSNR) in objective evaluation methods, this paper designed a questionnaire and published it on the Internet to discuss the relationship between SSIM and PSNR and study their relationship with subjective quality evaluation methods. We collected 88 volunteers from all walks of life and invited them to fill out a questionnaire designed by us. In addition, this paper also proposes a new subjective evaluation method. Inspired by the dual-stimulus loss classification method, we design a new scoring mechanism. The results show that SSIM and PSNR have high correlation, and the degree of their linear correlation fusion is different from subjective evaluation method.