Fine-grained hierarchical crop type classification from integrated hyperspectral EnMAP data and multispectral sentinel-2 time series: A large-scale dataset and dual-stream transformer method
- 1Jockey Club STEM Lab of Quantitative Remote Sensing, Department of Geography, The University of Hong Kong, Hong Kong, China
- 2Department of Aerospace Intelligent Science and Technology, School of Astronautics, Beihang University, Beijing, China
Abstract
Fine-grained crop type classification serves as the fundamental basis for large-scale crop mapping and plays a vital role in ensuring food security. It requires the simultaneous capture of both phenological dynamics (obtained from multi-temporal satellite data like Sentinel-2) and subtle spectral variations (demanding nanometer-scale spectral resolution from hyperspectral imagery). Research combining these two modalities remains scarce currently due to challenges in hyperspectral data acquisition and crop type annotation costs. To address these issues, we construct a hierarchical hyperspectral crop dataset (H 2 Crop) by integrating 30 m-resolution EnMAP hyperspectral data with Sentinel-2 time series. With over one million annotated field parcels organized in a four-tier crop taxonomy, H 2 Crop establishes a vital benchmark for fine-grained agricultural crop classification and hyperspectral image processing. We propose a dual-stream Transformer architecture that synergistically processes these modalities. It coordinates two specialized pathways: a spectral-spatial Transformer extracts fine-grained signatures from hyperspectral EnMAP data, while a temporal Swin Transformer extracts crop growth patterns from Sentinel-2 time series. The designed hierarchical classification head with hierarchical fusion then simultaneously delivers multi-level crop type classification across all taxonomic tiers. Experiments demonstrate that adding hyperspectral EnMAP data to Sentinel-2 time series yields a 4.2% average F1-score improvement (peaking at 6.3%). Extensive comparisons also confirm our method’s higher accuracy over existing deep learning approaches for crop type classification and the consistent benefits of hyperspectral data across varying temporal windows and crop change scenarios.
Related
- An Integrated Framework Coupling Hierarchical Hyperspectral i‐trait Screening, Gene Identification, and Functional Validation for Genetic Dissection of Salt Tolerance in Cotton
- Combining 3D-multispectral and hyperspectral imaging to identify environmental stress treatments imposed during plant growth.
- Multi-scale vision transformer model with UAV-based hyperspectral imagery for early identification of pine wilt disease infected trees
- Fine-scale grassland land-cover classification using UAV RGB and multispectral imagery with staged feature selection
- MetaMamba: Meta-learning with mamba for few-shot vegetation species classification using UAV-based hyperspectral imagery in desert rangeland
BibTeX
@article{li2026finegrained,
title = {{Fine-grained hierarchical crop type classification from integrated hyperspectral EnMAP data and multispectral sentinel-2 time series: A large-scale dataset and dual-stream transformer method}},
author = {Wenyuan Li and Shunlin Liang and Yuxiang Zhang and Liqin Liu and Keyan Chen and Yongzhe Chen and Han Ma and Jianglei Xu and Yichuan Ma and Shikang Guan and Zhenwei Shi},
journal = {Remote Sensing of Environment},
year = {2026},
month = jun,
doi = {10.1016/j.rse.2026.115525},
url = {https://doi.org/10.1016/j.rse.2026.115525},
note = {CC BY},
}
RIS
TY - JOUR
TI - Fine-grained hierarchical crop type classification from integrated hyperspectral EnMAP data and multispectral sentinel-2 time series: A large-scale dataset and dual-stream transformer method
AU - Wenyuan Li
AU - Shunlin Liang
AU - Yuxiang Zhang
AU - Liqin Liu
AU - Keyan Chen
AU - Yongzhe Chen
AU - Han Ma
AU - Jianglei Xu
AU - Yichuan Ma
AU - Shikang Guan
AU - Zhenwei Shi
PY - 2026
DA - 2026/06/08
JO - Remote Sensing of Environment
DO - 10.1016/j.rse.2026.115525
UR - https://doi.org/10.1016/j.rse.2026.115525
AB - Fine-grained crop type classification serves as the fundamental basis for large-scale crop mapping and plays a vital role in ensuring food security. It requires the simultaneous capture of both phenological dynamics (obtained from multi-temporal satellite data like Sentinel-2) and subtle spectral variations (demanding nanometer-scale spectral resolution from hyperspectral imagery). Research combining these two modalities remains scarce currently due to challenges in hyperspectral data acquisition and crop type annotation costs. To address these issues, we construct a hierarchical hyperspectral crop dataset (H 2 Crop) by integrating 30 m-resolution EnMAP hyperspectral data with Sentinel-2 time series. With over one million annotated field parcels organized in a four-tier crop taxonomy, H 2 Crop establishes a vital benchmark for fine-grained agricultural crop classification and hyperspectral image processing. We propose a dual-stream Transformer architecture that synergistically processes these modalities. It coordinates two specialized pathways: a spectral-spatial Transformer extracts fine-grained signatures from hyperspectral EnMAP data, while a temporal Swin Transformer extracts crop growth patterns from Sentinel-2 time series. The designed hierarchical classification head with hierarchical fusion then simultaneously delivers multi-level crop type classification across all taxonomic tiers. Experiments demonstrate that adding hyperspectral EnMAP data to Sentinel-2 time series yields a 4.2% average F1-score improvement (peaking at 6.3%). Extensive comparisons also confirm our method’s higher accuracy over existing deep learning approaches for crop type classification and the consistent benefits of hyperspectral data across varying temporal windows and crop change scenarios.
SN - 0034-4257
SN - 1879-0704
ER -