Deep Learning of Radiative Atmospheric Transfer with an Autoencoder

As electro-optical energy from the sun propagates through the atmosphere it is affected by radiative transfer effects including absorption, emission, and scattering. Modeling these affects is essential for scientific remote sensing measurements of the earth and atmosphere. For example, hyperspectral imagery is a form of digital imagery collected with many, often hundreds, of wavelengths of light in pixel. The amount of light measured at the sensor is the result of emitted sunlight, atmospheric radiative transfer, and the reflectance off the materials on the ground, all of which vary per wavelength resulting from multiple physical phenomena. Therefore measurements of the ground spectra or atmospheric constituents requires separating these different contributions per wavelength. In this paper, we create an autoencoder similar to denoising autoencoders treating the atmospheric affects as 'noise' and ground reflectance as truth per spectrum. We generate hundreds of thousands of training samples by taking random samples of spectra from laboratory measurements and adding atmospheric affects using physics-based modelling via MODTRAN (http://modtran.spectral.com/modtran\_home) by varying atmospheric inputs. This process ideally could create an autoencoder that would separate atmospheric effects and ground reflectance in hyperspectral imagery, a process called atmospheric compensation which is difficult and time-consuming requiring a combination of heuristic approximations, estimates of physical quantities, and physical modelling. While the accuracy of our method is not as good as other methods in the field, this an important first step in applying the growing field of deep learning of physical principles to atmospheric compensation in hyperspectral imagery and remote sensing.

Neural Network Learning of Chemical Bond Representations in Spectral Indices and Features

In this paper we investigate neural networks for classification in hyperspectral imaging with a focus on connecting the architecture of the network with the physics of the sensing and materials present. Spectroscopy is the process of measuring light reflected or emitted by a material as a function wavelength. Molecular bonds present in the material have vibrational frequencies which affect the amount of light measured at each wavelength. Thus the measured spectrum contains information about the particular chemical constituents and types of bonds. For example, chlorophyll reflects more light in the near-IR rage (800-900nm) than in the red (625-675nm) range, and this difference can be measured using a normalized vegetation difference index (NDVI), which is commonly used to detect vegetation presence, health, and type in imagery collected at these wavelengths. In this paper we show that the weights in a Neural Network trained on different vegetation classes learn to measure this difference in reflectance. We then show that a Neural Network trained on a more complex set of ten different polymer materials will learn spectral 'features' evident in the weights for the network, and these features can be used to reliably distinguish between the different types of polymers. Examination of the weights provides a human-interpretable understanding of the network.

Classifying Crop Types using Gaussian Bayesian Models and Neural Networks on GHISACONUS USGS data from NASA Hyperspectral Satellite Imagery

Hyperspectral Imagining is a type of digital imaging in which each pixel contains typically hundreds of wavelengths of light providing spectroscopic information about the materials present in the pixel. In this paper we provide classification methods for determining crop type in the USGS GHISACONUS data, which contains around 7,000 pixel spectra from the five major U.S. agricultural crops (winter wheat, rice, corn, soybeans, and cotton) collected by the NASA Hyperion satellite, and includes the spectrum, geolocation, crop type, and stage of growth for each pixel. We apply standard LDA and QDA as well as Bayesian custom versions that compute the joint probability of crop type and stage, and then the marginal probability for crop type, outperforming the non-Bayesian methods. We also test a single layer neural network with dropout on the data, which performs comparable to LDA and QDA but not as well as the Bayesian methods.