KEYWORDS: Image compression, Chemical analysis, Wavelets, Hyperspectral imaging, Image filtering, Data modeling, Digital micromirror devices, Statistical analysis, Sensors, Image resolution, Multiresolution signal processing, Long wavelength infrared, Chemical detection
In this paper we derive two algorithms for estimating concentrations of a known chemical compound from compressed measurements of a hyperspectral image (HSI). It is assumed that each resolved pixel in a scene contains a chemical of known spectral signature, at an unknown concentration. The problem is to estimate the concentration directly from the compressed measurements. Estimated concentrations are either displayed or used as detection scores in a threshold test for presence or absence of chemical. In the first algorithm we use matched filtering and ℓ1 regularization to extract an image of concentrations, directly from compressed data. In the second we model the image of concentrations in a fixed-resolution subspace of the 2D Haar wavelet domain, estimate its parameters in this space, and reconstruct the image of concentrations at a macro-pixel resolution. We evaluate our algorithms by applying them to several long-wave infrared (LWIR) HSI data sets, either synthetically generated or recorded by Physical Sciences Inc. Synthetically-generated data is compressed with a mathematically-defined linear compressor; real HSI data is compressed with PSI’s Digital Micromirror Device (DMD), which is a physical implementation of a mathematically-defined compressor; Fabry-Perot data is raw HSI data recorded by PSI, which is then compressed with a mathematically-defined compressor. We demonstrate for these data sets that estimating concentrations through matched filtering and ℓ1 inversion of compressed measurements yields detection performance that is as good as previously proposed methods that first reconstruct a hyperspectral data cube from compressed data, and then estimate or detect chemical concentrations. The proposed methods save on memory and computation. We demonstrate that detection performance is maintained when resolving concentration maps at a lower resolution, so long as the resolution is not too low.
An environmentally hardened compressive sensing hyperspectral imager (CS-HSI) operating in the long wave infrared (LWIR) has been developed for low-cost, standoff, wide area early warning of chemical vapor plumes. The CS-HSI employs a single-pixel architecture achieving an order of magnitude cost reduction relative to conventional HSI systems and a favorable pixel fill factor for standoff chemical plume imaging. A low-cost digital micromirror device modified for use in the LWIR is used to spatially encode the image of the scene; a Fabry-Perot tunable filter in conjunction with a single element mercury cadmium telluride photo-detector is used to spectrally resolve the spatially compressed data. A CS processing module reconstructs the spatially compressed spectral data, where both the measurement and sparsity basis functions are tailored to the CS-HSI hardware and chemical plume imaging. An automated target recognition algorithm is applied to the reconstructed hyperspectral data employing a variant of the adaptive cosine estimator for detection of chemical plumes in cluttered and dynamic backgrounds. The approach also offers the capability to generate detection products in compressed space with no CS reconstruction. This detection in transform space can be performed with a computationally lighter minimum variance distortionless response algorithm, resulting in a bandwidth advantage that supports efficient search and confirm modes of operation.
A novel multi-path extinction detector (M-PED) is being developed for point detection, identification and quantification of vapor phase chemicals. M-PED functions by pairing a broadband long-wave infrared (LWIR) quantum cascade laser with a novel sample cell, designed to simultaneously measure chemical absorption at multiple pathlengths and wavelengths. The pathlength samples are angularly separated in one dimension, such that a diffraction grating can be used to measure wavelength data in the orthogonal dimension using a compact, low-cost microbolometer array. The resulting data matrix is fit to Beer’s Law in two dimensions to accurately quantify chemical concentration while rejecting common mode noise (e.g. laser amplitude noise). The design, characterization and a capability demonstration of the advanced prototype sensor are presented.
A compressive sensing hyperspectral imaging (CS-HSI) platform has been developed for low-cost, standoff, wide area Early Warning of chemical vapor plumes. The sensor, operating in the longwave infrared (LWIR) spectral range with a single-pixel architecture, simultaneously addresses two practical shortcomings of LWIR chemical plume imaging platforms: (1) the single pixel architecture enables an order of magnitude cost reduction relative to HSI sensors employing a cooled focal plane array or high-speed gimbaled scanner, and (2) the inherent imaging modality achieves a favorable pixel fill factor and associated probability of detection for relevant chemical threats relative to single pixel scanned sensors. The CS-HSI employs a low-cost digital micromirror device modified for use in the LWIR spectral range to spatially encode an image of the scene. An LWIR spectrometer employing a tunable Fabry-Perot filter and a mercury cadmium telluride single element photo-detector spectrally resolves the spatially integrated image while mitigating instrument radiance. A CS processing module reconstructs the spatially compressed hyperspectral image where the measurement and sparsity basis functions are specifically tailored to the CS-HSI hardware and chemical plume imaging. An automated target recognition algorithm is applied to the reconstructed hyperspectral data employing a variant of the Adaptive Cosine Estimator for the detection of the chemical plumes in cluttered and dynamic backgrounds. The development, characterization, and a series of capability demonstrations of a prototype CS-HSI sensor are presented. Capability demonstrations include chemical plume imaging of R-134 at mission-relevant concentration pathlength product levels in a laboratory setting.
Sampling is a fundamental aspect of any implementation of compressive sensing. Typically, the choice of sampling method is guided by the reconstruction basis. However, this approach can be problematic with respect to certain hardware constraints and is not responsive to domain-specific context. We propose a method for defining an order for a sampling basis that is optimal with respect to capturing variance in data, thus allowing for meaningful sensing at any desired level of compression. We focus on the Walsh-Hadamard sampling basis for its relevance to hardware constraints, but our approach applies to any sampling basis of interest. We illustrate the effectiveness of our method on the Physical Sciences Inc. Fabry-Pérot interferometer sensor multispectral dataset, the Johns Hopkins Applied Physics Lab FTIR-based longwave infrared sensor hyperspectral dataset, and a Colorado State University Swiss Ranger depth image dataset. The spectral datasets consist of simulant experiments, including releases of chemicals such as GAA and SF6. We combine our sampling and reconstruction with the adaptive coherence estimator (ACE) and bulk coherence for chemical detection and we incorporate an algorithmic threshold for ACE values to determine the presence or absence of a chemical. We compare results across sampling methods in this context. We have successful chemical detection at a compression rate of 90%. For all three datasets, we compare our sampling approach to standard orderings of sampling basis such as random, sequency, and an analog of sequency that we term `frequency.' In one instance, the peak signal to noise ratio was improved by over 30% across a test set of depth images.
An active, standoff, all-phase chemical detection capability has been developed under IARPA’s SILMARILS program. The detection platform utilizes reflectance spectroscopy in the longwave infrared coupled with an automated detection algorithm that implements physics-based reflectance models for planar chemical films, particulate in the solid and liquid phase, and vapors. Target chemicals include chemical warfare agents, toxic industrial chemicals, and explosives. The platform employs broadband Fabry-Perot quantum cascade lasers with a spectrally selective detector to interrogate target surfaces at tens of meter standoff. A statistical F-test in a noise whitened space is used for detection and discrimination over a large target spectral library in high clutter environments.
The capability is described with an emphasis on the physical reflectance models used to predict spectral reflectivity signatures as a function of surface contaminant presentation and loading. Developmental test results from a breadboard version of the detector platform are presented. Specifically, solid and liquid surface contaminants were detected and identified from a library of 325 compounds down to 30 μg/cm2 surface loading at a 5 m standoff. Vapor detection was demonstrated via topographic backscatter.
Advances towards the development of a longwave infrared quantum cascade laser (QCL) based standoff and proximal surface contaminant detection platform are presented with emphasis on developmental test results. The detection platform utilizes reflectance spectroscopy with application to optically thick and thin materials in film and particulate forms including solid and liquid phase chemical warfare agents, toxic industrial chemicals and materials, and explosives. The platform employs an ensemble of broadband Fabry-Perot QCLs with a spectrally selective detector to interrogate target surfaces at 1 to 10s of m standoff. A version of a Subspace Adaptive Cosine Estimator is used for detection and discrimination in high clutter environments. Through speckle reduction, a noise equivalent reflectivity of 0.1% was demonstrated enabling detection limits approaching 0.1 μg/cm2 for optically thin films and 2% fill factor for optically thick particulates.
The design, build, and validation of a breadboard version of the QCL-based surface contaminant detector are summarized. Results from developmental testing of contaminated substrates in standoff (5 m range) and proximal (~1 m range) configurations are presented. The test substrates were prepared by the government and Physical Sciences, Inc. and include solid and liquid contaminants at varying surface loadings. Future improvements including an expanded spectral range are discussed.
Progress towards the development of a longwave infrared quantum cascade laser (QLC) based standoff surface contaminant detection platform is presented. The detection platform utilizes reflectance spectroscopy with application to optically thick and thin materials including solid and liquid phase chemical warfare agents, toxic industrial chemicals and materials, and explosives. The platform employs an ensemble of broadband QCLs with a spectrally selective detector to interrogate target surfaces at 10s of m standoff. A version of the Adaptive Cosine Estimator (ACE) featuring class based screening is used for detection and discrimination in high clutter environments. Detection limits approaching 0.1 μg/cm2 are projected through speckle reduction methods enabling detector noise limited performance.
The design, build, and validation of a breadboard version of the QCL-based surface contaminant detector are discussed. Functional test results specific to the QCL illuminator are presented with specific emphasis on speckle reduction.
Sensor technologies capable of detecting low vapor pressure liquid surface contaminants, as well as solids, in a noncontact fashion while on-the-move continues to be an important need for the U.S. Army. In this paper, we discuss the development of a long-wave infrared (LWIR, 8-10.5 μm) spatial heterodyne spectrometer coupled with an LWIR illuminator and an automated detection algorithm for detection of surface contaminants from a moving vehicle. The system is designed to detect surface contaminants by repetitively collecting LWIR reflectance spectra of the ground. Detection and identification of surface contaminants is based on spectral correlation of the measured LWIR ground reflectance spectra with high fidelity library spectra and the system’s cumulative binary detection response from the sampled ground. We present the concepts of the detection algorithm through a discussion of the system signal model. In addition, we present reflectance spectra of surfaces contaminated with a liquid CWA simulant, triethyl phosphate (TEP), and a solid simulant, acetaminophen acquired while the sensor was stationary and on-the-move. Surfaces included CARC painted steel, asphalt, concrete, and sand. The data collected was analyzed to determine the probability of detecting 800 μm diameter contaminant particles at a 0.5 g/m2 areal density with the SHSCAD traversing a surface.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.