Underwater Image Denoising and Semantic Segmentation

The goal of this research is to denoise underwater image quality and precisely segment underwater objects. Imaging underwater is hampered by noise, blur, and color distortion. Denoising and semantic segmentation steps make up our method. To minimize noise and improve image quality, we use 2D-Variational Mode Decomposition, white balance, and LIME method. For accurate object segmentation, sophisticated deep learning architectures are employed. Extensive testing reveals considerable advancements in precise segmentation and denoising. Our method has potential in surveillance, marine biology, and underwater robots.