四川轻化工大学学报（自然科学版）2025，Vol.38Issue(6)：17-28,12.DOI:10.11863/j.suse.2025.06.03
高光谱与多光谱图像融合技术综述
A Review of Hyperspectral and Multispectral Image Fusion Technique
摘要
Abstract
Hyperspectral and multispectral image fusion(HS-MS fusion),as a crucial technique for enhancing the spatial resolution of remote sensing(RS)images,has emerged as a focal point in multi-source remote sensing image fusion(MSRS fusion)research.Firstly,the concept of MSRS fusion is introduced.Secondly,the recent advances in HS-MS fusion from the perspectives of fusion levels,architectures,and methodologies are comprehensively reviewed.The fusion levels are classified into three categories including pixel-level,feature-level and decision-level;the fusion structures are summarized into three types containing serial,parallel and hybrid;the fusion methods are comprehensively categorized into multiple categories such as pan-sharpening-based and supervised learning-based methods,and the differences and connections between traditional methods,deep learning methods,and their combined fusion methods are analyzed.Finally,the applications of HS-MS fusion in agricultural monitoring and climate change research are discussed,the problems and challenges faced in multi-source information mining and scenario-specific personalized requirements are summarized,and the targeted improvement solutions are provided.Moreover,prospects for future development are outlined.关键词
遥感图像融合/高光谱图像/多光谱图像/多源信息融合Key words
remote sensing image fusion/hyperspectral image/multispectral image/multi-source information fusion分类
信息技术与安全科学引用本文复制引用
张星月,陈明举,胡潇,李森远,宋晓飞,饶杰..高光谱与多光谱图像融合技术综述[J].四川轻化工大学学报（自然科学版）,2025,38(6):17-28,12.基金项目
四川省自然科学基金项目(2025ZNSFSC0477 （）
2024NSFSC2042) （）
四川轻化工大学研究生创新基金资助项目(Y2024297) （Y2024297）
