Advancements in Image Dehazing: Enhanced Techniques with Non-uniform Atmospheric Light, Improved Dark Channel Prior, and Combined Window Filter

This chapter delves into the challenges posed by haze in images, affecting both ground-based and remote sensing photographs. It explores the limitations of existing dehazing methods, which often focus on a single type of haze or suffer from color distortions and excessive contrast. The proposed method introduces a novel technique that leverages non-uniform atmospheric light estimation and a combined window filter to enhance image dehazing. The chapter provides a detailed overview of the methodology, including image acquisition, contrast enhancement using CLAHE, and the estimation of atmospheric light through patch-based analysis. It also discusses the role of the guided filter in smoothing and the normalization of images to improve dehazing accuracy. The dark channel prior is enhanced with a combined window filter to preserve edges and remove textures, resulting in a more accurate transmission map. The chapter concludes with experimental results demonstrating the effectiveness of the proposed method across various types of haze in both ground-captured and remote sensing images, supported by quantitative assessments such as PSNR and SSIM.