Information-efficient spectral imaging sensor

• US 6,504,943 B1
• Filed: 06/29/1999
• Issued: 01/07/2003
• Est. Priority Date: 07/20/1998
• Status: Expired due to Term

• Current Assignee: National Technology & Engineering Solutions Of Sandia LLC (Honeywell International Inc.)
• Original Assignee: Sandia Corporation (Martin Marietta Corporation)
• Inventors: Descour, Michael R., Stallard, Brian R., Grotbeck, Carter L., Boye, Clinton A., Sweatt, William C., Gentry, Stephen M.
• Primary Examiner(s): Boudreau, Leo
• Assistant Examiner(s): Choobin, Barry
• Application Number:
  US09/345,604
• Time in Patent Office: 1,288 Days
• Field of Search: 382/103, 382/100, 356/310, 348/269, 348/169
• US Class Current: 382/103
• CPC Class Codes: G02B 27/1086 operating by diffraction only
• CPC Class Codes: G02B 27/123 The splitting element being a lens or a system of lenses, including arrays and surfaces with refractive power
• CPC Class Codes: G02B 27/126 The splitting element being a prism or prismatic array, including systems based on total internal reflection
• CPC Class Codes: G02B 27/145 having sequential partially reflecting surfaces
• CPC Class Codes: G02B 27/28 for polarising used in stereoscopes G02B30/25
• CPC Class Codes: G02B 27/4205 having a diffractive optical element [DOE] contributing to image formation, e.g. whereby modulation transfer function MTF or optical aberrations are relevant
• CPC Class Codes: G02B 27/4294 in multispectral systems, e.g. UV and visible
• CPC Class Codes: G06T 2207/10152 Varying illumination
• CPC Class Codes: G06T 2207/30108 Industrial image inspection
• CPC Class Codes: G06T 5/50 using two or more images, e.g. averaging or subtraction
• CPC Class Codes: G06T 7/0008 checking presence/absence
• CPC Class Codes: G06V 20/13 Satellite images

Abstract

A programmable optical filter for use in multispectral and hyperspectral imaging. The filter splits the light collected by an optical telescope into two channels for each of the pixels in a row in a scanned image, one channel to handle the positive elements of a spectral basis filter and one for the negative elements of the spectral basis filter. Each channel for each pixel disperses its light into n spectral bins, with the light in each bin being attenuated in accordance with the value of the associated positive or negative element of the spectral basis vector. The spectral basis vector is constructed so that its positive elements emphasize the presence of a target and its negative elements emphasize the presence of the constituents of the background of the imaged scene. The attenuated light in the channels is re-imaged onto separate detectors for each pixel and then the signals from the detectors are combined to give an indication of the presence or not of the target in each pixel of the scanned scene. This system provides for a very efficient optical determination of the presence of the target, as opposed to the very data intensive data manipulations that are required in conventional hyperspectral imaging systems.

Claims

1. A method emphasizing a first aspect of a set of spectral data from an imaged scene with a line-scanning multispectral sensor having m-pixels per row, with n spectral bins sensed per pixel in a row, the method comprising:

utilizing a previously-created first spectral basis vector having n elements to identify the first aspect, some of which elements have negative values and the rest having positive values, said vector being derived from a training set of a multiplicity of n element spectra that includes at least one spectrum of the first aspect and at least one spectrum of the constituents of the background materials in the scene;

to collecting light from the imaged scene and presenting it to the sensor;

attenuating the light in the affected spectral bins of each pixel of a row in a first channel of the sensor based on the value of the respective positive value elements of the first spectral basis vector;

attenuating the light in the affected spectral bins of each pixel of that row in a second channel of the sensor based on the value of the respective negative value elements of the first spectral basis vector;

imaging the modulated light in the first and second channels onto respective first and second linear arrays of detectors, each detector in each array corresponding to a pixel in that row, to provide first and second detector signals; and

combining the first and second detector signals to provide an indication of the presence or not of the first aspect in the scanned pixels.

2. The method of claim 1 additionally comprising the step of:

utilizing a previously-created second spectral basis vector having n elements to emphasize a second aspect of the imaged scene, some of which elements have negative values and the rest having positive values, from a training set of a multiplicity of n element spectra to provide an indication of the presence or not of the second aspect in the scanned pixels based on the attenuations to the collected light caused by the second spectral basis vector in parallel with the utilization of the first spectral basis vector.

3. The method of claim 2 additionally comprising the steps of:

splitting the collected light between the first spectral basis vector and the second spectral basis vector;

attenuating the light in the affected spectral basis bins of each pixel of a row in a third channel of the sensor based on the value of the respective positive value elements of the spectral basis vector for the second aspect;

attenuating the light in the affected spectral basis bins of each pixel of the row in a fourth channel of the sensor based on the value of the respective negative value elements of the spectral basis vector for the second aspect;

imaging the modulated light in the third and fourth channels onto respective third and fourth detector arrays to provide third and fourth detector signals; and

combining the third and fourth detector signals to provide an indication of the presence or not of the second aspect in the scanned pixels.

4. The method of claim 1 wherein the attenuation of the light in the first and second channels by the first spectral basis vector is responsive to a controller that is responsive to commands from a remote location.

5. The method of claim 1 wherein the light from the scene is presented to the sensor by a slit oriented parallel to the row of pixels.

6. The method of claim 5 wherein the scene is scanned in a direction perpendicular to the long axis of the slit.

7. The method of claim 6 wherein each detector in each array is a time delay and integrate (TDI) detector array that electronically accumulates the signals from its detectors in synchronization with the motion of the scene across such detectors.

8. A line scanning, multispectral sensor having at least two optical channels to emphasize at least one aspect of a scanned scene comprising:

a slit to form a linear image of m pixels in a row;

means to direct the light from the slit for each pixel into first and second channels of light with the light in each channel being dispersed into n spectral bins and to present this light to individual optical attenuators for each spectral bin in each channel, wherein the levels of attenuation, if any, in each attenuator in the first channel correspond to the positive elements in a first spectral basis vector that emphasizes a first aspect of the scanned scene and the levels of attenuation, if any, in each attenuator in the second channel correspond to the negative elements in a spectral basis vector;

detector arrays for each channel with a detector for each spectral bin therein to provide an electrical signal corresponding to the strength of the light leaving the attenuators in each spectral bin in each channel; and

means to combine the signals from the two detector arrays to emphasize the first aspect of the scanned scene;

wherein the first spectral basis vector is derived from a training set of a multiplicity of n element spectra that includes at least one spectrum of the first aspect and at least one spectrum of the constituents of the background materials in the scene to emphasize the first aspect of the scanned scene.

9. The sensor of claim 8 wherein the means to direct the light into first and second channels with the light in each channel being dispersed into n spectral bins include polarizing beamsplitters and micromirrors, and gratings and prisms, respectively.

10. The sensor of claim 8 further including a controller that adjusts the levels of attenuation in the individual attenuators.

11. The sensor of claim 8 wherein the detectors in the detector arrays comprise time delay and integrate (TDI) arrays that electronically accumulate the signals from its detectors in synchronization with the motion of the scene across such detectors.

12. A line scanning, multispectral sensor having at least two optical channels to emphasize at least one aspect of a scanned scene comprising:

optical elements suitable to image a scene onto a slit;

a slit to form a linear image that is separable into m pixels in a row;

a second optical element suitable to recollimate the light from the slit onto a dispersing element;

a dispersing element to split the light from the slit into n spectral bins that are disposed normal to the axis of the slit;

a third optical element suitable to converge the light from the dispersing element and image the n spectral images of the slit onto n rows of a micromirror array;

a micromirror array comprising a linear array of n individual micromirror assemblies with each assembly corresponding to a single spectral bin such that each assembly is controlled to reflect the light from for its spectral bin into at least two different positions, with a first and second position corresponding to first and second channels, wherein the light in a spectral bin is reflected into the first channel if the corresponding element in a first n element spectral basis vector is positive and the light in that spectral bin is reflected into the second channel if the corresponding element in the first spectral basis vector is negative;

p1 a fourth optical element suitable to reimage the light in the first channel from the micromirror array onto a first m-element detector array;

a fifth optical element suitable to reimage the light in the second channel from the micromirror array onto a second m-element detector array; and

means to combine the signals from the first and second detector arrays to emphasize the first aspect in each of the pixels of the scanned scene, wherein the first spectral basis vector is derived from a training set of a multiplicity of n element spectra that includes at least one spectrum of the first aspect and at least one spectrum of the constituents of the background materials in the scene to emphasize the first aspect of the scanned scene.

13. The sensor of claim 12 wherein the individual micromirror assemblies include at least two micromirrors in each assembly.

14. The sensor of claim 12 wherein the dispersing element is a grating or a prism.

15. The sensor of claim 12 wherein the detectors in the first and second m-element detector arrays comprise time delay and integrate (TDI) arrays that electronically accumulate the signals from its detectors in synchronization with the motion of the scene across such detectors.

16. A line scanning, multispectral sensor having at least two optical channels to emphasize at least one aspect of a scanned scene comprising:

optical elements suitable to image a scene onto a slit;

a slit to form a linear image that is separable into m pixels in a row;

a second optical element suitable to recollimate the light from the slit onto a first polarizer;

a dispersing element to separate the light into n spectral bins;

a third optical element suitable to recollimate and refocus the light from the dispersing element onto an array of liquid crystal spatial light modulators (SLMs);

an array of SLMs with one SLM for each spectral bin, with each SLM being responsive to signals from a controller with respect to the amount of rotation imparted to the polarization vector of the light passing through that SLM corresponding to the positive and negative elements of a first spectral basis vector;

a fourth optical element suitable to recollimate the light from the array of SLMs onto an analyzing polarizer;

an analyzing polarizer that directs unrotated light from the array of SLMs into a first optical channel and directs rotated light from the array of SLMS into a second optical channel;

a first optical channel that includes a fifth optical element suitable to focus light onto a first detector array;

a second optical channel that includes a sixth optical element suitable to focus light onto a second detector array; and

means to combine the signals from the first and second detector arrays in order to emphasize a first aspect in the pixels of the scanned scene, wherein the first spectral basis vector is derived from a training set of a multiplicity of n element spectra that includes at least one spectrum of the first aspect and at least one spectrum of the constituents of the background materials in the scene to emphasize the first aspect of the scanned scene.

17. The sensor of claim 16 wherein the dispersing element is a grating or a prism.

18. The sensor of claim 16 wherein the detectors in the first and second detector arrays comprise time delay and integrate (TDI) arrays that electronically accumulate the signals from its detectors in synchronization with the motion of the scene across such detectors.