Mako airborne thermal infrared imaging spectrometer: performance update

Hall, Jeffrey L.; Boucher, Richard H.; Buckland, Kerry N.; Gutierrez, David J.; Keim, E. R.; Tratt, David M.; Warren, David W.

doi:10.1117/12.2239245

Cited by 13 publications

(12 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This section first reviews the atmospheric measurement data and corresponding forward-modeled TUD vectors used for dimension reduction. Metrics for comparing the performance of each technique are also reviewed, with a focus on incorporating properties from the simplified RT model in Equation (1). A unique data augmentation scheme is also discussed to increase the number of TUD samples for model fitting.…”

Section: Methodsmentioning

confidence: 99%

“…Next-generation hyperspectral imagers continue to improve in both spatial and spectral resolution with increasingly lower noise-equivalent spectral radiance (NESR) values, presenting unique opportunities in efficiently characterizing pixel materials [1]. A pixel in a hyperspectral image can be represented as a vector across all spectral channels, producing a three-dimensional data cube for an entire image, width by height by spectral channel [2].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Fast and Effective Techniques for LWIR Radiative Transfer Modeling: A Dimension-Reduction Approach

2019

View full text Add to dashboard Cite

The increasing spatial and spectral resolution of hyperspectral imagers yields detailed spectroscopy measurements from both space-based and airborne platforms. These detailed measurements allow for material classification, with many recent advancements from the fields of machine learning and deep learning. In many scenarios, the hyperspectral image must first be corrected or compensated for atmospheric effects. RT computations can provide LUT to support these corrections. This research investigates a dimension-reduction approach using machine learning methods to create an effective sensor-specific lwir RT model. The utility of this approach is investigated emulating the Mako lwir hyperspectral sensor ( Δ λ ≃ 0 . 044 m , Δ ν ˜ ≃ 3 . 9 c m - 1 ). This study employs physics-based metrics and loss functions to identify promising dimension-reduction techniques and reduce at-sensor radiance reconstruction error. The derived RT model shows an overall RMSE of less than 1 K across reflective to emissive grey-body emissivity profiles.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Fast and Effective Techniques for LWIR Radiative Transfer Modeling: A Dimension-Reduction Approach

2019

View full text Add to dashboard Cite

show abstract

“…The LWIR Mako hyperspectral sensor is a high-performance, airborne sensor imaging across 7.8 − 13.4 µm into 128 spectral channels with a noise-equivalent temperature difference of 0.02K at 10 µm and 300 K [1,30]. The high-resolution LBLRTM generated TUD vectors (11,513 spectral channels) are downsampled according to the Mako instrument line shape creating representative TUD vectors for this sensor.…”

Section: Datamentioning

confidence: 99%

“…Next generation hyperspectral imagers continue to improve in both spatial and spectral resolution with increasingly lower noise-equivalent spectral radiance (NESR) values, presenting unique challenges and opportunities in efficiently characterizing pixel materials [1]. A pixel in a hyperspectral image can be represented as a vector across all spectral channels, producing a three dimensional data cube for an entire image, width by height by spectral channel [2].…”

Section: Introductionmentioning

confidence: 99%

“…• Employing machine learning techniques which: (1) are computationally faster than traditional RT calculation methods; (2) reduce the dimension of both the TUD and atmospheric measurement vectors; (3) produce the desirable latent-space-similarity property such that small deviations in the low-dimension latent space result in small deviations in the high-dimension TUD • Developing a data augmentation method using PCA and Gaussian Mixture Models (GMMs) on real atmospheric measurements that lead to improved model training and generalizability • Improving machine learning model training by introducing a physics-based loss function which encourages better fit models than traditional loss functions based on mean squared error • Demonstrating an effective Autoencoder (AE) pre-training strategy that leverages the local-similarity properties of the latent space to reproduce TUDs from atmospheric conditions Together, these contributions form the basis of a novel method for efficient and effective RT modeling, using a small number of parameters, which is applicable for atmospheric compensation across diverse remote sensing scenarios.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Fast and Effective Techniques for LWIR Radiative Transfer Modeling: A Dimension Reduction Approach

Westing¹,

Borghetti²,

Gross³

2019

Preprint

View full text Add to dashboard Cite

The increasing spatial and spectral resolution of hyperspectral imagers yields detailed spectroscopy measurements from both space-based and airborne platforms. Machine learning algorithms have achieved state-of-the-art material classification performance on benchmark hyperspectral data sets; however, these techniques often do not consider varying atmospheric conditions experienced in a real-world detection scenario. To reduce the impact of atmospheric effects in the at-sensor signal, atmospheric compensation must be performed. Radiative Transfer (RT) modeling can generate high-fidelity atmospheric estimates at detailed spectral resolutions, but is often too time-consuming for real-time detection scenarios. This research utilizes machine learning methods to perform dimension reduction on the transmittance, upwelling radiance, and downwelling radiance (TUD) data to create high accuracy atmospheric estimates with lower computational cost than RT modeling. The utility of this approach is investigated using the instrument line shape for the Mako long-wave infrared hyperspectral sensor. This study employs physics-based metrics and loss functions to identify promising dimension reduction techniques. As a result, TUD vectors can be produced in real-time allowing for atmospheric compensation across diverse remote sensing scenarios.

show abstract