Underwater Image Enhancement Based on Depth and Light Attenuation Estimation
Lianjun Zhang, Tingna Liu, Qichao Shi, Fen Chen · IET Image Processing · 2025
ABSTRACT Light attenuation and complex water environments seriously deteriorate underwater imaging quality. Current underwater image restoration algorithms cannot handle low‐quality colour‐distorted images in aquatic environments. This study proposed a novel underwater image processing algorithm based on a light attenuation estimation model and a depth estimation network. First, a pseudo‐depth map strategy was proposed to train the underwater image depth estimation network to realise underwater image depth estimation. Second, the attenuation coefficient of the current image was estimated based on the background light using a light attenuation model. Finally, the images were restored using an underwater imaging model. The proposed algorithm is superior to state‐of‐the‐art underwater image processing algorithms regarding subjective and objective qualities.