现代电子技术2024，Vol.47Issue(19)：87-93,7.DOI:10.16652/j.issn.1004-373x.2024.19.014
基于细节增强框架的图像超分辨率算法
Image super-resolution algorithm based on detail-enhancing framework
摘要
Abstract
In the technology of the reference-based image super-resolution(Ref-SR),the high-resolution(HR)reference images are imported into the single-input image super-resolution(SISR)approach,and the long-standing ill-posed issue of the very method has been alleviated with the assistance of texture transferred from reference images.Although the significant improvement in quantitative and qualitative results has verified the superiority of Ref-SR methods,the presence of misalignment before texture transfer indicates room for further performance improvement.Existing methods tend to neglect the significance of details in the context of comparison,so they fail to leverage the information contained within low-resolution(LR)images.In view of the above,a reference-based image super-resolution method on the basis of detail-enhancing framework(DEF),which introduces the diffusion model to generate and enhance the underlying detail in LR images,is proposed.If corresponding parts are present in the reference images,the proposed method can facilitate rigorous alignment.In cases where the reference images lack corresponding parts,the details in the LR images are still enhanced,and the influence of the wrong texture transferred from the reference images is avoided.Extensive experiments demonstrate that the proposed method achieves superior visual results while maintaining comparable numerical results in comparison with the existing methods.关键词
图像超分辨率/参考图像/不适定问题/相关对齐/纹理转移/扩散模型Key words
image super-resolution/reference image/ill-posed issue/correlation alignment/texture transfer/diffusion model分类
信息技术与安全科学引用本文复制引用
王子涵,袁晓兵..基于细节增强框架的图像超分辨率算法[J].现代电子技术,2024,47(19):87-93,7.基金项目
微系统技术实验室基金项目(6142804230103) （6142804230103）
