Dark Channel Prior-Based Single-Image Dehazing Using Type-2 Fuzzy Sets for Edge Enhancement in Dehazed Images

The process of image dehazing is known to have a significant impact in the field of computer vision. It has been found to have practical applications in various real-world scenarios, such as autonomous driving, surveillance, and the enhancement of outdoor images. This study tackles the challenge of image dehazing by introducing an approach that combines the dark channel prior (DCP) technique with type-2 fuzzy set theory. The objective of the proposed methodology is to improve the image quality by utilizing the I-Haze and O-Haze datasets, which encompass a compilation of hazy images captured under diverse environmental circumstances. In order to assess the efficacy of our approach, we employ various quality assessment metrics including Peak Signal to Noise Ratio (PSNR), Structural Similarity Index (SSIM), Lightness Order Error (LOE), and Naturalness Image Quality Evaluator (NIQE). Our research contributes to advancing image dehazing techniques and their practical applications, across domains.