Physical Degradation-Guided Deep Unfolding for Blind Cross-Sensor Spectral Super-Resolution

Hyperspectral imaging provides rich spectral information for quantitative remote sensing, yet hyperspectral sensors remain costly and thus unavailable in many UAV deployments. Spectral super-resolution (SSR) seeks to reconstruct hyperspectral images (HSIs) from multispectral images (MSIs). Many model-guided SSR methods assume a fixed and known spectral response function (SRF), whereas the effective spectral transformation between different HSI and MSI sensors is often unavailable. This paper proposes a physical degradation-guided deep unfolding network, termed PGU-Net, to address blind cross-sensor SSR by jointly estimating the HSI and a learnable spectral transformation function (STF). PGU-Net unfolds the alternating estimation of the HSI and STF into a multi-stage network, where closed-form solvers enforce data fidelity and learnable proximal networks model the two priors. Experiments on simulated and remote sensing datasets under different spectral degradations demonstrate accurate STF estimation and improved HSI reconstruction. Spatial three-fold validation on the real Headwall–DJI sensor pair further verifies PGU-Net under blind cross-sensor conditions.