Global Attention Super-Resolution Algorithm for Nature Image Edge Enhancement

Single-image super-resolution (SR) has long been a research hotspot in computer vision, playing crucial role practical applications such as medical imaging, public security and remote sensing imagery. However, all currently available methods focus on reconstructing texture details, resulting blurred edges incomplete structures the reconstructed images. To address this problem, an edge-enhancement-based global attention image network (EGAN) combining channel- self-attention mechanisms is proposed for modeling hierarchical features intra-layer multiple dimensions. Specifically, channel contrast-aware (CCA) module learns correlations between feature channels enhances contrast maps richer edge structures. The cyclic shift window multi-head (CS-MSA) captures long-range dependencies layered more valuable information network. Experiments are conducted five benchmark datasets × 2, 3 4 SR. experimental results show that SR, our improves average PSNR by 0.12 dB, 0.19 dB over RCAN, HAN NLSN, respectively, can reconstruct clear complete structure.