Akademik Veri Yönetim Sistemi
INTERNATIONAL JOURNAL OF REMOTE SENSING, 2026 (SCI-Expanded, Scopus)
The objective of this study is to enhance the 10 m (B7-B12) and 20 m (B13-B19) spatial resolution bands of the Jilin-1 GP02 satellite imagery through fusion with the 5 m (B0-B6) bands, thereby producing high-accuracy 5 m hyperspectral (HS) bands and improving the usability of this dataset for diverse remote sensing applications. A series of experiments were conducted using different data configurations and state-of-the-art fusion algorithms to assess the impact of spectral compatibility, scale ratio and algorithmic design on fusion performance. Among all configurations, Experiment 2.c, which fused the B7-B19 bands with a spectrally compatible synthetic panchromatic (PAN) image generated by averaging the B1-B6 bands, achieved the best quantitative and qualitative results. These results confirm that employing a synthetic PAN image closely aligned with the spectral characteristics of the target HS data effectively minimizes spectral distortion and enhances spatial fidelity compared with conventional PAN- or MS-based fusion schemes. Algorithmic evaluations further demonstrated that physically and contextually informed models such as C-MTF-GLP-CBD, PRACS, and BDSD-PC exhibited robust and consistent performance across different environments, while HS-MS fusion approaches (Experiments 3.a and 3.b) were found to be less effective. This study contributes to the literature by presenting one of the first comprehensive evaluations of high-resolution HS fusion using Jilin-1 GP02 data, which is gaining increasing importance in global Earth observation. The results establish a methodological framework for generating high-resolution HS products from Jilin-1 GP02 imagery and provide valuable insights to guide future research and technological development in data fusion for advanced Earth observation applications.