Research on Adaptive Dehazing Algorithm Based on Image Illuminance

This paper proposes an image illuminance-adaptive defogging algorithm, aiming to address the issues of image distortion and inadequate consideration of diverse lighting environments inherent in traditional methods, thereby enhancing the algorithm's universal applicability. The algorithm initially refines the image through superpixel segmentation, subsequently detecting point light sources precisely and estimating the overall illumination conditions, providing a precise environmental basis for the defogging process. Based on the acquired illuminance information, defogging parameters are adjusted to achieve defogging tailored to varying lighting conditions. To further elevate the visual quality of defogged images, this algorithm integrates the Contrast Limited Adaptive Histogram Equalization (CLAHE) technique with the point light source detection mechanism, enhancing image contrast while maintaining a natural appearance and minimizing noise amplification. Experimental results demonstrate that, compared to conventional methods, the proposed algorithm achieves notable improvements in key evaluation metrics such as Structural Similarity Index Measure (SSIM) and Peak Signal-to-Noise Ratio (PSNR) across foggy images under diverse lighting conditions, effectively enhancing the quality of defogged images.