CARNet: Context-Aware Residual Learning for JPEG-LS Compressed Remote Sensing Image Restoration

JPEG-LS (a lossless (LS) compression standard developed by the Joint Photographic Expert Group) compressed image restoration is a significant problem in remote sensing applications. It faces following two challenges: first, bridging small pixel-value gaps from wide numerical ranges; and second, removing banding artifacts condition of lacking available context information. As far as we know, there currently no research dealing with above issues. Hence, develop this initial line work on restoration. We propose novel CNN model called CARNet. Its core idea context-aware residual learning mechanism. Specifically, it realizes for accurate adopting scale-invariant baseline. enables large receptive fields artifact removal through scheme. Additionally, eases information flow among stages utilizing prior-guided feature-fusion Alternatively, design R IQA models to provide better performance assessment our study gradient priors artifacts. Furthermore, prepare new dataset images supplement existing benchmark data. Experiments show that method sets state-of-the-art