DCT Underwater Image Enhancement Based on Attenuation Analysis.

Underwater images often suffer from color distortion, reduced contrast, and blurred details due to the selective absorption and scattering of light by water, which limits the performance of underwater visual tasks. To address these issues, this paper proposes an underwater image enhancement method that integrates multi-channel attenuation analysis and discrete cosine transform (DCT). First, the color statistics of an in situ-captured underwater image are mapped to those of a reference image selected from a well-illuminated natural image dataset with standard color distribution; no pristine underwater image is required. This mapping yields a color transfer image, i.e., an intermediate color-corrected result obtained via statistical matching. Subsequently, this image is fused with an attenuation weight map and the original input to produce the final color-corrected result. Secondly, taking advantage of the median's resistance to extreme value interference and the Sigmoid function's flexible control of gray-scale transformation, the gray-scale range is adjusted in different regions through nonlinear mapping to achieve global contrast balance. Finally, considering the visual system's sensitivity to high-frequency details, a saliency map is extracted using Gabor filtering, and the frequency characteristics are analyzed through block DCT transformation. Adaptive gain is applied to high-frequency details to enhance them. Experiments were conducted on the UIEB, EUVP, and LSUI datasets and compared with existing methods. Through qualitative and quantitative analysis, it was verified that the proposed algorithm not only effectively enhances underwater images but also significantly improves image clarity.