Underwater Image Enhancement Algorithm Based on Illuminance Domain Analysis

Abstract

Aiming at the problem of poor image quality in conventional underwater vision system, especially in the case of poor water quality, a new underwater image enhancement algorithm based on illuminance domain analysis and improved DCP algorithm and Retinex algorithm is proposed in this paper. This enhancement algorithm estimates the background illumination of the water body of the imported underwater image, compares the estimated value with the set critical illumination value, and adopts the corresponding improved algorithm to obtain the enhanced image, which effectively solves the imaging problem in the poor water quality environment. Uciqe method is used to evaluate the quality of the enhanced image, and more accurate and objective experimental results are obtained. The enhanced algorithm can be applied to underwater fishing, dike inspection, nuclear facilities inspection, treasure hunting, archaeology, fish assessment, ship management and customs inspection.