Image Enhancement by Defogging utilizing Scene Specific Dark Channel Prior
Agnesh Chandra Yadav, Samprit Bose, Maheshkumar H. Kolekar · 2024
Adverse weather conditions such as haze and fog significantly impact the visibility and quality of outdoor images, leading to errors in various computer vision systems. Traditional defogging methods, often fall short in terms of efficiency and accuracy. In this study, an approach is presented to address these challenges. The presented method involves estimating the approximate depth map of a hazy image, from which the scene-specific dark channel and transmission are derived. Post-processing includes histogram normalization, further enhancing the overall image quality. By incorporating these steps, the technique aims to surpass the limitations of conventional defogging algorithms, offering improved efficiency and accuracy in restoring the quality of individual images affected by adverse weather conditions. The outcomes from the experiments demonstrated a noteworthy enhancement in both visible and quantitative aspects compared to other leading techniques.