Local Background Estimation for Improved Gas Plume Identification in Hyperspectral Images

TL;DR

Urban LWIR hyperspectral gas plume identification is hindered by heterogeneous backgrounds, which can degrade global whitening. The paper introduces LEBEAUS, a local background estimation framework combining Watershed segmentation, iBATE iterative background estimation, and robust segment-distance metrics (TAL and TED) to tailor background spectra under each background material. It demonstrates improved plume whitening and higher deep-learning identification confidence on both simulated SF6/NH3 plumes and real LA basin plumes, with hyperparameters showing cross-plume robustness. This approach offers a practical preprocessing enhancement for gas detection in complex scenes, potentially enabling more reliable hyperspectral plume analysis across diverse plumes and gases.

Abstract

Deep learning identification models have shown promise for identifying gas plumes in Longwave IR hyperspectral images of urban scenes, particularly when a large library of gases are being considered. Because many gases have similar spectral signatures, it is important to properly estimate the signal from a detected plume. Typically, a scene's global mean spectrum and covariance matrix are estimated to whiten the plume's signal, which removes the background's signature from the gas signature. However, urban scenes can have many different background materials that are spatially and spectrally heterogeneous. This can lead to poor identification performance when the global background estimate is not representative of a given local background material. We use image segmentation, along with an iterative background estimation algorithm, to create local estimates for the various background materials that reside underneath a gas plume. Our method outperforms global background estimation on a set of simulated and real gas plumes. This method shows promise in increasing deep learning identification confidence, while being simple and easy to tune when considering diverse plumes.

Figures (2)

• Figure 1: Plume ROIs shown in red over false-colored context images. First row are the simulated plumes, second row are the real plumes.
• Figure 2: Hyperparameter search slice plots. The orange x marks the optimal hyperparameters, and the horizontal line marks the baseline performance from global whitening. The first two rows are the simulated gases where lower MSE is preferred. The last two rows are the real plumes, where higher confidence is preferred.