UAV Multispectral to Hyperspectral Reconstruction Based on the RTM and Deep Learning

The dataset is organized as follows: data/real/Contains experimentally measured data, including: UAV hyperspectral observations ASD field spectrometer measurementsThese data are used as real validation and fine-tuning data in our experiments. data/lut/Contains training data generated using a Look-Up Table (LUT) approach.The LUT is generated based on: Wheat-related biophysical parameters from our experiment Parameter ranges reported in relevant literatureThese data are used for LUT-based pretraining of the model. data/ancillary/Contains soil spectral data, used as ancillary input for LUT generation and simulation. data/sensor/Contains the spectral response functions (SRFs) of the UAV sensor, used to match simulated spectra with UAV observations. All data provided in this Zenodo record are raw data only.Source code for data generation, model training, and evaluation is maintained separately on GitHub https://github.com/windying123/M2H-SWIR.git.