Multispectral to Hyperspectral using Pretrained Foundational model
This work addresses the challenge of precise greenhouse gas detection for atmospheric monitoring by combining the strengths of hyperspectral and multispectral imaging, though it appears incremental as it builds on existing transformer methods and datasets.
The study tackled the problem of limited spatial and temporal coverage in hyperspectral imaging for greenhouse gas monitoring by reconstructing hyperspectral data from multispectral inputs using transformer models, achieving results through pretraining on EnMAP and EMIT datasets and fine-tuning on aligned image pairs.
Hyperspectral imaging provides detailed spectral information, offering significant potential for monitoring greenhouse gases like CH4 and NO2. However, its application is constrained by limited spatial coverage and infrequent revisit times. In contrast, multispectral imaging delivers broader spatial and temporal coverage but lacks the spectral granularity required for precise GHG detection. To address these challenges, this study proposes Spectral and Spatial-Spectral transformer models that reconstruct hyperspectral data from multispectral inputs. The models in this paper are pretrained on EnMAP and EMIT datasets and fine-tuned on spatio-temporally aligned (Sentinel-2, EnMAP) and (HLS-S30, EMIT) image pairs respectively. Our model has the potential to enhance atmospheric monitoring by combining the strengths of hyperspectral and multispectral imaging systems.