计算机与数字工程2025，Vol.53Issue(1)：209-213,5.DOI:10.3969/j.issn.1672-9722.2025.01.038
一种针对图像融合的Dense-FCNN算法研究
Research on Dense-FCNN Algorithm for Image Fusion
摘要
Abstract
High spatial and hyperspectral images can bring more comprehensive information to computer vision processing,es-pecially in remote sensing image fusion.In recent years,the remote sensing image fusion method based on convolution network has brought a breakthrough in the field of remote sensing image fusion.Aiming at the quality problem of remote sensing image fusion,a kind of Dense-FCNN network is proposed.The network combines the characteristics of DenseNet network and FusionCNN.Based on the idea of making full use of the characteristics,the DenseBlock module is added to the different structures of FusionCNN,and a comparative analysis is made.The results show that the improved dense FCNN network has improved the fusion quality,and the best model effect has increased by 3.4%.Moreover,the fusion image can better retain the spatial information of Pan image and the spectral information of MS image,which has practical application value for obtaining more comprehensive remote sensing image in-formation.关键词
融合质量/卷积网络/图像融合/空间信息/光谱信息Key words
fusion quality/convolutional network/image fusion/spatial information/spectral information分类
信息技术与安全科学引用本文复制引用
郑迦馨,罗银辉,吴岳洲,王宇..一种针对图像融合的Dense-FCNN算法研究[J].计算机与数字工程,2025,53(1):209-213,5.基金项目
国家自然科学基金项目"大数据驱动的飞行训练智能评估理论与方法"(编号:U2033213) （编号:U2033213）
国家重点研发计划项目(编号:2021YFF0603904) （编号:2021YFF0603904）
中央高校基本科研业务费基金项目(编号:ZJ2022-004)资助. （编号:ZJ2022-004）
