Abstract:• We develop a novel encoder-decoder architecture with an attention-based skip fusion integrated into autonomous underwater robot for underwater image enhancement. • We propose a multi-scale deformable convolutional network, which acquires deformable local receptive fields at different scales and enriches the diversity of feature representation. • We adopt a parallel attention block for spatial and channel dimensions to focus on the interest regions of the input feature map. • Our MsDCANet achieves state-of-the-art performance on the recent benchmark in terms of visual effects and object metrics, which is conducive to integrating underwater robots and applications in aquaculture. Underwater images suffer from poor visibility quality, which results from selective attenuation and scattering. These interdependence phenomena together cause image degradation, failing autonomous underwater robots to recognize image contents. To address those problems, we propose a multi-scale deformable convolution network with an attention mechanism (MsDCANet) to enhance the quality of underwater images. The proposed model is generally implemented by an encoder-decoder architecture. Concretely, we first propose a multi-scale deformable convolutional network, which acquires deformable local receptive fields at different scales and enriches the diversity of feature representation. Coupled with spatial and channel attention mechanisms, the task-oriented high-level representations acquired from input feature maps, and the most significant features are highlighted. Considering the contributions of the encoder and decoder functions on the final enhancement task, an attention-based skip fusion scheme is designed to improve the original skip connection, which aims to reconstruct higher quality underwater images. Finally, the proposed model is optimized by a multi-task loss function, including pixel loss and perceptual loss. Experiments on underwater images captured under diverse scenes demonstrate that the proposed model produces visually pleasing results and has strong generalization ability, even significantly outperforming several state-of-the-art methods. Besides, our approach can also improve the performance of vision tasks and be applied to aquaculture development.

Multi-scale Iterative Network for Underwater Image Restoration

Multi-scale convolution underwater image restoration network

Underwater image restoration with multi-scale shallow feature extraction and detail enhancement network

Color correction and restoration based on multi-scale recursive network for underwater optical image

MDSCN: multiscale depthwise separable convolutional network for underwater graphics restoration

Autonomous Underwater Robot for Underwater Image Enhancement Via Multi-Scale Deformable Convolution Network with Attention Mechanism

Underwater Image Enhancement Network Based on Dual Layers Regression

A novel underwater image restoration method based on decomposition network and physical imaging model

Enhancing Underwater Images Via Asymmetric Multi-Scale Invertible Networks

A Multi-Scale Feature Extraction Network for Marine Underwater Image Enhancement

A Cascaded Multimodule Image Enhancement Framework for Underwater Visual Perception

Synergistic Multiscale Detail Refinement via Intrinsic Supervision for Underwater Image Enhancement

Joint Iterative Color Correction and Dehazing for Underwater Image Enhancement

Two-Stage Underwater Image Restoration Algorithm Based on Physical Model and Causal Intervention

Adaptive Deep Learning Network with Multi-Scale and Multi-Dimensional Features for Underwater Image Enhancement

Degradation-aware and Color-Corrected Network for Underwater Image Enhancement

Underwater Image Enhancement via Learning Water Type Desensitized Representations

Underwater Image Enhancement via Dehazing and Color Restoration

Underwater image restoration based on dual information modulation network

Underwater Image Processing System For Image Enhancement And Restoration

Jointly Adversarial Network to Wavelength Compensation and Dehazing of Underwater Images