Detection by active spatially and spectrally structured sensing and learning (DAS4L)

• US 10,295,462 B1
• Filed: 02/28/2017
• Issued: 05/21/2019
• Est. Priority Date: 03/02/2016
• Status: Active Grant

• Current Assignee: HRL Laboratories LLC (The Boeing Co.)
• Original Assignee: HRL Laboratories LLC (The Boeing Co.)
• Inventors: Yap, Daniel, Owechko, Yuri, Kremer, Richard M., Rao, Shankar R.
• Primary Examiner(s): Kim, Kiho
• Application Number:
  US15/445,782
• Time in Patent Office: 812 Days
• CPC Class Codes: G01N 2021/945 Liquid or solid deposits of macroscopic size on surfaces, e.g. drops, films, or clustered contaminants dust particles and microscopic contaminants in G01N21/94
• CPC Class Codes: G01N 21/27 using photo-electric detection G01N21/31 takes precedence; circuits for computing concentration logarithmic circuits G06G7/24; photometric circuits in general G01J
• CPC Class Codes: G01N 21/31 Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry G01N21/72 takes precedence
• CPC Class Codes: G01N 21/35 using infrared light G01N21/39 takes precedence
• CPC Class Codes: G01N 2201/1293 resolving multicomponent spectra
• CPC Class Codes: G01N 33/227 Explosives, e.g. combustive properties thereof detecting explosives in air G01N33/0057

Abstract

A sensor system including a spectrometer with a light source having a plurality of selectable wavelengths, a controller for controlling the sensor system, for selecting wavelengths of illumination light produced by the light source, and for controlling the light source to illuminate a spatial location, a photodetector aligned to detect light received from the spatial location, a blind demixer coupled to the photodetector for separating received spectra in the detected light into a set of sample spectra associated with different demixed or partially demixed chemical components, a memory having a plurality of stored reference spectra, a non-blind demixer coupled to the blind demixer and to the memory for non-blind demixing of the sample spectra using the reference spectra, and a classifier coupled to the non-blind demixer for classifying the set of demixed sample spectra into chemical components using the reference spectra.

Claims

1. A sensor system comprising:

a spectrometer with a light source having a plurality of selectable wavelengths;

a controller for controlling the sensor system, for selecting wavelengths of illumination light produced by the light source, and for controlling the light source to illuminate a spatial location;

a photodetector aligned to detect light received from the spatial location;

a blind demixer coupled to the photodetector for separating received spectra in the detected light into a set of sample spectra associated with different chemical components;

a background spectrum remover coupled to the photodetector and to blind demixer, the background spectrum remover capable of removing from the received spectra spectra known to be associated with spectral clutter;

a memory having a plurality of stored reference spectra;

a non-blind demixer coupled to the blind demixer and to the memory for non-blind demixing of the sample spectra using the reference spectra; and

a classifier coupled to the non-blind demixer for classifying the set of demixed sample spectra into chemical components using the reference spectra.

2. The sensor system of claim 1:

wherein the controller controls the sensor system to probe an area with illumination light with coarsely spaced wavelengths to distinguish between regions of interest containing residues and residue free background regions;

wherein the classifier is coupled to the controller and controls the sensor system to probe the regions of interest with illumination light having wavelengths having a finer spacing than the coarsely spaced wavelengths to determine the chemical components in the regions of interest.

3. The sensor system of claim 1 wherein the blind demixer comprises an independent component analysis (ICA) algorithm.

4. The sensor system of claim 1 wherein the classifier comprises a sparse representation-based classification (SRC) algorithm.

5. The sensor system of claim 1 further comprising:

a confidence-weighted chemical-identifier coupled to the classifier for determining confidence weights for the identified chemical components.

6. The sensor system of claim 5 further comprising:

a known material comparer coupled to memory and to the confidence-weighted chemical-identifier for comparing the identified chemical components to known materials;

wherein the known material comparer makes a final determination of the chemical component if the chemical component is a known material; and

wherein the known material comparer updates the reference spectra in the memory if the chemical component is not a known material.

7. The sensor system of claim 1 wherein the spectral clutter is typically obtained in a measurement of a mixture of a chemical residue and a background material.

8. The sensor system of claim 1:

wherein the controller controls the sensor system to probe an area with coarsely spaced wavelengths to distinguish between regions of interest that contain residues and background regions that are residue free;

wherein the controller controls the sensor system to probe the regions of interest and the background regions with wavelengths having a finer spacing than the coarsely spaced wavelengths to determine the chemical components in the regions of interest and the background regions; and

wherein the spectra determined to be spectral clutter in the background regions are removed from the received spectra of the regions of interest by the background spectrum remover.

9. The sensor system of claim 1:

wherein the reference spectra comprise one or more of measurements of pure chemical residues, measurements of mixtures containing several chemical species, measurements of substrate materials, and measurements of spectral deformations resulting from various measurement conditions, including angle of illumination light, residue thickness, and texture.

10. A sensor system comprising:

a spectrometer with a light source having a plurality of selectable wavelengths;

a controller for controlling the sensor system, for selecting wavelengths of illumination light produced by the light source, and for controlling the light source to illuminate a spatial location;

a photodetector aligned to detect light received from the spatial location;

a blind demixer coupled to the photodetector for separating received spectra in the detected light into a set of sample spectra associated with different chemical components;

a memory having a plurality of stored reference spectra;

a non-blind demixer coupled to the blind demixer and to the memory for non-blind demixing of the sample spectra using the reference spectra; and

a classifier coupled to the non-blind demixer for classifying the set of demixed sample spectra into chemical components using the reference spectra;

wherein the memory further comprises a stored hierarchically structured decision tree having a plurality of branches each coupled to a branch higher or lower in the hierarchically structured decision tree or to a termination leaf;

wherein a respective branch higher in the hierarchically structured decision tree has characteristics associated with lower branches or termination leafs coupled to the respective branch higher in the hierarchically structured decision tree; and

wherein each termination leaf is associated with a respective reference spectrum.

11. A sensor system comprising:

a spectrometer with a light source having a plurality of selectable wavelengths;

a controller for controlling the sensor system, for selecting wavelengths of illumination light produced by the light source, and for controlling the light source to illuminate a spatial location;

a photodetector aligned to detect light received from the spatial location;

a blind demixer coupled to the photodetector for separating received spectra in the detected light into a set of sample spectra associated with different chemical components;

a memory having a plurality of stored reference spectra;

a non-blind demixer coupled to the blind demixer and to the memory for non-blind demixing of the sample spectra using the reference spectra; and

a classifier coupled to the non-blind demixer for classifying the set of demixed sample spectra into chemical components using the reference spectra;

wherein the memory further comprises a stored hierarchically clustered spectral library;

wherein a cluster in the hierarchically clustered spectral library is formed by measuring each reference spectra and constructing an N-dimensional vector associated with each reference spectrum, wherein a length of a vector in each of N-dimensions of an N-dimensional vector is a value of the spectral intensity at a measured wavenumber; and

wherein reference spectra having N-dimensional vectors whose Euclidean distance is not greater than a threshold are grouped in the same cluster.

12. A sensor system comprising:

a spectrometer with a light source having a plurality of selectable wavelengths;

a controller for controlling the sensor system, for selecting wavelengths of illumination light produced by the light source, and for controlling the light source to illuminate a spatial location;

a photodetector aligned to detect light received from the spatial location;

a blind demixer coupled to the photodetector for separating received spectra in the detected light into a set of sample spectra associated with different chemical components;

a memory having a plurality of stored reference spectra;

a non-blind demixer coupled to the blind demixer and to the memory for non-blind demixing of the sample spectra using the reference spectra; and

a classifier coupled to the non-blind demixer for classifying the set of demixed sample spectra into chemical components using the reference spectra;

a plurality of transmitter units, wherein each of the transmitter units simultaneously transmits a light beam having one of the plurality of selectable wavelengths; and

an optical system coupled to the plurality of transmitter units and the photodetector, wherein the optical system directs the light beam from each of the transmitter units onto the illuminated spot, and wherein the optical system collects light from the illuminated spot and directs the light to the photodetector.

13. A method for detecting and identifying chemical components, the method comprising:

transmitting to a spatial location light having a first plurality of simultaneously occurring wavelengths;

measuring an intensity of back-scattered light from the first set of the plurality of simultaneously occurring wavelengths;

removing from the measured intensity of the back-scattered light from the first plurality of simultaneously occurring wavelengths spectra known to be associated with spectral clutter;

using the measured intensity of the back-scattered light to make an intermediate identification of the chemical components at the spatial location;

using the intermediate identification to determine a second a plurality of simultaneously occurring wavelengths for illuminating the spatial location, wherein the second plurality of simultaneously occurring wavelengths has a finer wavelength spacing than the first plurality of simultaneously occurring wavelengths;

transmitting to the spatial location light having the second plurality of simultaneously occurring wavelengths;

measuring an intensity of second back-scattered light from the second set of the plurality of simultaneously occurring wavelengths; and

using the measured intensity of the second back-scattered light to make an identification of the chemical components at the spatial location.

14. The method of claim 13 further comprising:

using a plurality of reference spectra to determine the second set of the plurality of wavelengths.

15. The method of claim 14 wherein using the measured intensity of the second back-scattered light to make an identification of the chemical components comprises:

non-blind demixing and classification using the plurality of reference spectra;

wherein the reference spectra comprise one or more of measurements of pure chemical residues, measurements of mixtures containing several chemical species, measurements of substrate materials, and measurements of spectral deformations resulting from various measurement conditions, including angle of illumination light, residue thickness, and texture.

16. The method of claim 14 further comprising:

comparing the identified chemical components to known materials;

making a final determination of the chemical component if the chemical component is a known material; and

updating the reference spectra if the chemical component is not a known material.

17. The method of claim 13 wherein using the measured intensity of the back-scattered light to make an intermediate identification of the chemical components comprises:

blind demixing.

18. The method of claim 13 wherein using the measured intensity of the second back-scattered light to make an identification of the chemical components comprises:

non-blind demixing and classification using a sparse representation-based classification (SRC) algorithm.

19. The method of claim 13 further comprising:

controlling a sensor system to probe an area with illumination light with coarsely spaced wavelengths to distinguish between regions of interest containing residues and residue free background regions;

controlling the sensor system to probe the regions of interest and the background regions with wavelengths having a finer spacing than the coarsely spaced wavelengths to determine the chemical components in the regions of interest and the background regions; and

removing spectra determined to be spectral clutter in the background regions from the received spectra of the regions of interest.

20. The method of claim 13 further comprising:

determining confidence weights for the identified chemical components.

21. The method of claim 13 wherein the spectral clutter is typically obtained in a measurement of a mixture of a chemical residue and a background material.

22. A method for detecting and identifying chemical components, the method comprising:

transmitting to a spatial location light having a first plurality of simultaneously occurring wavelengths;

measuring an intensity of back-scattered light from the first plurality of simultaneously occurring wavelengths;

using the measured intensity of the back-scattered light to make an intermediate identification of the chemical components at the spatial location;

using the intermediate identification to determine a second plurality of simultaneously occurring wavelengths for illuminating the spatial location, wherein the second plurality of simultaneously occurring wavelengths has a finer wavelength spacing than the first plurality of simultaneously occurring wavelengths;

transmitting to the spatial location light having the second plurality of simultaneously occurring wavelengths;

measuring an intensity of second back-scattered light from the second plurality of simultaneously occurring wavelengths; and

using the measured intensity of the second back-scattered light to make an identification of the chemical components at the spatial location;

using a hierarchically structured decision tree having a plurality of branches each coupled to a branch higher or lower in the hierarchically structured decision tree or to a termination leaf;

wherein a respective branch higher in the hierarchically structured decision tree has characteristics associated with lower branches or termination leafs coupled to the respective branch higher in the hierarchically structured decision tree; and

wherein each termination leaf is associated with a respective reference spectrum.

23. A method for detecting and identifying chemical components, the method comprising:

transmitting to a spatial location light having a first plurality of simultaneously occurring wavelengths;

measuring an intensity of back-scattered light from the first plurality of simultaneously occurring wavelengths;

using the measured intensity of the back-scattered light to make an intermediate identification of the chemical components at the spatial location;

using the intermediate identification to determine a second plurality of simultaneously occurring wavelengths for illuminating the spatial location, wherein the second plurality of simultaneously occurring wavelengths has a finer wavelength spacing than the first plurality of simultaneously occurring wavelengths;

measuring an intensity of second back-scattered light from the second plurality of simultaneously occurring wavelengths; and

using the measured intensity of the second back-scattered light to make an identification of the chemical components at the spatial location

using a hierarchically clustered spectral library;

wherein a cluster in the hierarchically clustered spectral library is formed by measuring a plurality of reference spectra and constructing an N-dimensional vector associated with each reference spectrum, wherein a length of a vector in each of N-dimensions of an N-dimensional vector is a value of the spectral intensity at a measured wavenumber; and

wherein reference spectra having N-dimensional vectors whose Euclidean distance is not greater than a threshold are grouped in the same cluster.