<p>Due to the presence of suspended particles in the water, underwater images are susceptible to light scattering and occlusion, resulting in a significant decrease in image visibility and contrast, which causes image blur and loss of detail, and increases the difficulty of image processing and object recognition. To address these problems, this paper proposes an underwater image enhancement algorithm based on classification adaptive color correction and a dual parallel branch optimization network, called AMUE. Specifically, we first propose to dynamically adjust the color balance according to an adaptive factor to adapt to different underwater scenes. By analyzing the brightness mean, standard deviation, dynamic range, and saturation of the image, we divide the image into different color distortion levels and dynamically adjust the color balance factor accordingly. In addition, we design a dual-parallel branch network to optimize image quality: one branch performs global color saturation enhancement and denoising, and the other branch improves image texture by enhancing local details. Finally, an adaptive weighted multi-task fusion strategy is adopted to dynamically adjust the fusion weights according to the gradient and structural features of the image to ensure that the image achieves the best balance in clarity, contrast, and naturalness. The results show that the algorithm can effectively adapt to a variety of underwater scenes and provide reliable support for underwater vision tasks.</p>