Image Restoration of Atmospheric Turbulence Degradation Based on Space-Channel Attention Adversarial Network
The randomness and complexity of atmospheric turbulence affect the spatial positioning of light waves during transmission,leading to image degradation in long-range imaging and reducing the ability of devices to extract image infor-mation.To address the issue of turbulence-degraded images,propose a spatial-channel attention adversarial network(SCA-GAN)for restoring turbulence-degraded images.The network progressively eliminates noise,geometric distortions,and blurring in degraded images.By applying spatial-channel attention,the model extracts features that capture both local fine details and global contextual information,while integrating inter-channel information to achieve content consistency in the restored images.Experimental results demonstrate the feasibility of the proposed method in the field of restoring turbulence-degraded images.
万方数据
维普
