Polarized pixelated filter array with reduced sensitivity to misalignment for polarimetric imaging

Patent metadata

Field	Value
Publication number	US-10168542-B2
Application number	US-201514836305-A
Country	US
Kind code	B2
Filing date	Aug 26, 2015
Priority date	Aug 26, 2015
Publication date	Jan 1, 2019
Grant date	Jan 1, 2019

Abstract

Polarized pixelated filter sub-array is reconfigured to reduce sensitivity to misalignment. The condition number increases more slowly than the standard polarized pixelated filter sub-array as the misalignment increases. In different embodiments, the filter sub-array is configured such that the condition number has a finite bound at ½ pixel misalignment. The angular values of the polarizer filter array are determined to minimize the sensitivity of the condition number of the data reduction matrix to misalignment. This can be achieved by selecting angular values that minimize the expected value of the condition number E(CN) over the range of misalignment.

First claim

We claim:

1. A method of fabrication of a polarized pixelated filter array to selectively polarize electromagnetic radiation for use with a pixelated detector having dimensions in the x and y directions for which sub-arrays of three or more pixelated detector pixels P are read out and processed to form a polarimetric image, said method comprising:
selecting a size of a filter sub-array having dimensions in the x and y directions to include three or more filter pixels Q where Q>=P;
defining a non-zero probability distribution p ma (x,y) that models the misalignment of the filter sub-array in the x and y directions relative to the pixelated detector as a random variable with a range of misalignment in the x and y directions;
defining a merit function including at least a first term W 1 *E(CN) where W 1 is a weight and E(CN) is an expected value of a condition number (CN) of a data reduction matrix for the filter sub-array over the range of misalignment;
evaluating the merit function for a plurality of candidate filter sub-arrays, each grouping of P pixels in a candidate filter sub-array including at least two pixels that impart a linear polarization of different angular values;
selecting the candidate filter sub-array that minimizes the merit function; and
fabricating the polarized pixelated filter array comprising one or more of the selected candidate filter sub-arrays for use with the pixelated detector.

2. The method of claim 1 , wherein each said grouping of P pixels in the filter sub-array includes at least one unpolarized pixel.

3. The method of claim 1 , wherein said pixelated detector pixels are read out and processed in M×M sub-arrays where M>=2 and said filter sub-array pixels are arranged in an N×N grid format in the x and y directions where N is >=M, each M×M grouping of pixels in the filter sub-array including at least two pixels that impart a linear polarization of a different angular value.

4. The method of claim 3 , wherein M=2.

5. The method of claim 4 , wherein N>2.

6. The method of claim 4 , wherein N=3, said 3×3 grid format comprising four 2×2 groupings of pixels in the filter sub-array, each said grouping including at least three pixels that impart linear polarizations of different angular values or at least two pixels that impart linear polarizations of different angular values and at least one unpolarized pixel.

7. The method of claim 3 , wherein each 2×2 grouping of pixels in the filter sub-array includes at least one unpolarized pixel.

8. The method of claim 7 , where N is an even integer.

9. The method of claim 1 , wherein said range of misalignment spans 0 to ½ pixel in the x or y directions.

10. The method of claim 1 , wherein sub-arrays of M×M pixelated detector pixels are read out and processed to form the polarimetric image, where M is at least 2, wherein the size of the filter sub-array is selected to include N×N pixels where N>=M.

11. The method of claim 10 , wherein M=2 and N>M.

12. The method of claim 11 , wherein at least one pixel in each 2×2 grouping is an unpolarized pixel.

13. The method of claim 12 , wherein the merit function includes a second term W 2 *(1/Tx) where W 2 is a weight and Tx is the transmittance of unpolarized light from a point source, further comprising selecting non-zero values of the weights W 1 and W 2 .

14. The method of claim 10 , wherein each 2×2 grouping of pixels in the filter sub-array include at least three pixels that impart linear polarizations of different angular values or at least two pixels that impart linear polarizations of different angular values and at least one unpolarized pixel.

15. The method of claim 1 , wherein the pixelated detector and filter sub-array pixels are of equal size and shape and laid out on a rectilinear grid format.

16. The method of claim 1 , wherein sub-arrays of 2×2 pixelated detector pixels are read out and processed to form the polarimetric image, wherein selecting the candidate filter sub-array comprises selecting a filter sub-array in which a plurality of pixels are arranged in a 4×4 grid format in the x and y directions, including 8 pixels that impart a linear polarization of a certain angular value and 8 unpolarized, U, pixels, wherein the relative angular values for the 4×4 grid format are:

U
70°
U
110°

215°
U
5°
U

U
140°
U
20°

200°
U
45°
U.

17. A method of fabrication of a polarized pixelated filter array to selective polarize electromagnetic radiation for use with a pixelated detector having dimensions in the x and y directions for which 2×2 sub-arrays of pixelated detector pixels are read out and processed to form a polarimetric image, said method comprising:
selecting a size of a filter sub-array having dimensions in the x and y directions to include N×N filter pixels where N>2;
defining a non-zero probability distribution p ma (x,y) that models the misalignment of the filter sub-array in the x and y directions relative to the pixelated detector as a random variable with a range of misalignment in the x and y directions;
defining a merit function including at least a first term W 1 *E(CN) where W 1 is a weight and E(CN) is an expected value of a condition number (CN) of a data reduction matrix for the filter sub-array over the range of misalignment;
evaluating the merit function for a plurality of candidate N×N filter sub-arrays, each 2×2 grouping of pixels in a candidate filter sub-array including at least three pixels that impart linear polarizations of different angular values or at least two pixels that impart linear polarizations of different angular values and at least one unpolarized pixel;
selecting the candidate filter sub-array that minimizes the merit function; and
fabricating the polarized pixelated filter array comprising one or more of the selected candidate filter sub-arrays for use with the pixelated detector.

18. The method of claim 17 , wherein at least one pixel in each 2×2 grouping in the selected candidate filter sub-array is an unpolarized pixel.

19. The method of claim 18 , wherein the merit function includes a second term W 2 *(1/Tx) where W 2 is a weight and Tx is the transmittance of unpolarized light from a point source, further comprising selecting non-zero values of the weights W 1 and W 2 .

Assignees

• Raytheon Co

Inventors

• Fest Eric C
• Leigh Jon E

Classifications

• G06T2207/10048

  Infrared image · CPC title

• G02B27/286 Primary

  for controlling or changing the state of polarisation, e.g. transforming one polarisation state into another (G02B5/3083 takes precedence; light guide coupling means utilising polarising elements G02B6/34) · CPC title

• G06T7/344

  involving models · CPC title

• G01J4/04

  Polarimeters using electric detection means (G01J4/02 takes precedence) · CPC title

• G02B5/201

  in the form of arrays · CPC title