Underwater Image Dehazing using Non-Local Prior Method and Air-Light Estimation

Water absorbs light differently than air and this can have a significant impact on the visual quality of underwater images. All of these factors combine to make underwater photography and videography resources challenging. Underwater image enhancement technology has numerous applications in various fields such as marine biology, oceanography and underwater exploration. Many underwater image dehazing algorithms have been introduced in the literature to reconstruct the pictographic information of the underwater images. But these methods still suffered from generating optimal results. In order to solve these problems, this study presents a competent underwater image dehazing method based on non-local prior and air-light estimation (NLP-ALE). First, non-local prior is used where it contains four major steps such as finding haze line, estimate initial transmission, regularization and initial dehazing operation. Next, we used the air-light estimation method, which depends on the prior for haze lines. This prior is based on the finding that lines in RGB space intersecting at the air-light can be used to describe the pixel values of a hazy image. To choose the location of the air-light, we apply the Hough transform in RGB space. The experimental findings show that the proposed algorithm produces results comparable to and even superior to those of recent state-of-the-art techniques along with the benefit of being appropriate for real time applications. The proposed research technique exhibits ideal performance in evaluation metrics such as PSNR and SSIM comparted to other competing methods.