Search for: All records

The constellation of Earth-observing satellites continuously collects measurements of scattered radiance, which must be transformed into geophysical parameters in order to answer fundamental scientific questions about the Earth. Retrieval of these parameters requires highly flexible, accurate, and fast forward and inverse radiative transfer models. Existing forward models used by the remote sensing community are typically accurate and fast, but sacrifice flexibility by assuming the atmosphere or ocean is composed of plane-parallel layers. Monte Carlo forward models can handle more complex scenarios such as 3D spatial heterogeneity, but are relatively slower. We propose looking to the computer graphics community for inspiration to improve the statistical efficiency of Monte Carlo forward models and explore new approaches to inverse models for remote sensing. In Part 2 of this work, we demonstrate that Monte Carlo forward models in computer graphics are capable of sufficient accuracy for remote sensing by extending Mitsuba 3, a forward and inverse modeling framework recently developed in the computer graphics community, to simulate simple atmosphere-ocean systems and show that our framework is capable of achieving error on par with codes currently used by the remote sensing community on benchmark results. 

The constellation of Earth-observing satellites continuously collects measurements of scattered radiance, which must be transformed into geophysical parameters in order to answer fundamental scientific questions about the Earth. Retrieval of these parameters requires highly flexible, accurate, and fast forward and inverse radiative transfer models. Existing forward models used by the remote sensing community are typically accurate and fast, but sacrifice flexibility by assuming the atmosphere or ocean is composed of plane-parallel layers. Monte Carlo forward models can handle more complex scenarios such as 3D spatial heterogeneity, but are relatively slower. We propose looking to the computer graphics community for inspiration to improve the statistical efficiency of Monte Carlo forward models and explore new approaches to inverse models for remote sensing. In Part 1 of this work, we examine the evolution of radiative transfer models in computer graphics and highlight recent advancements that have the potential to push forward models in remote sensing beyond their current periphery of realism. 

We propose a novel design for a do-it-yourself hyperspectral imaging system which operates by taking multiple photographs through tunable, polarization-induced, spectral filters. Prior approaches in this do-it-yourself arena achieve hyperspectral imaging by selecting from a discrete set of spectra baked into existing products. In contrast, our approach is capable of generating a continuous family of broadband transmission spectra by simple rotations of stacked polarizers and waveplates. This greatly expands the potential range of representable spectra from a fixed-dimensional to an arbitrary-dimensional space. We analyze the theoretical spectral gamut of our approach and demonstrate its viability for spectral surface reflectance reconstruction both in simulation and with a low-cost physical prototype. Our prototype demonstrates that our approach can achieve comparable quality to prior work at reduced cost, while the new design space holds ample opportunity for increased quality and flexibility with professional manufacturing.