Due to the differential absorptions of light waves by water, combined with the scattering and refraction effects caused by dissolved impurities, plankton, and suspended particles, underwater images suffer degradation issues such as color distortion, blurred details, low contrast, and insufficient illumination, etc. To address these challenges and limitations, a low illumination underwater image enhancement framework is designed in this paper. Color cast correction is carried out by dominant channel compensation. The dark region light brighten is implemented through logarithmic and linear transform, and we use dual-scale feature extraction technology to capture more important details which involve local details, textural information, and global structural shapes. The contrast enhancement is executed via the mean and standard deviation within the Saturation and Value channels of the HSV color space. Experimental comparison results on five different datasets with other eight advanced algorithms showed that the proposed algorithm average UCIQE and UIQM, AG and IE score is 0.5983 and 5.5118, 17.4066 and 7.6432 across all datasets, separately, and the second-best algorithm average scores are 0.5737 and 5.2923, respectively, which demonstrates that the our method exhibits superior color clarity and overall contrast, richer image gradation and texture features, while maintaining the strong robustness and generalization. The comparation result showed an improvement of 4.29% and 4.15%, 25.68% and 1.85%. These results demonstrate that our method delivers superior performance in terms of color preservation, clarity, and suitable overall contrast.