Image adjustment based on locally flat scenes

Patent metadata

Field	Value
Publication number	US-10467736-B2
Application number	US-201815940795-A
Country	US
Kind code	B2
Filing date	Mar 29, 2018
Priority date	Dec 2, 2014
Publication date	Nov 5, 2019
Grant date	Nov 5, 2019

Abstract

Imaging systems and methods are disclosed that use locally flat scenes to adjust image data. An imaging system includes an array of photodetectors configured to produce an array of intensity values corresponding to light intensity at the photodetectors. The imaging system can be configured to acquire a frame of intensity values, or an image frame, and analyze the image frame to determine if it is locally flat. If the image frame is locally flat, then that image data can be used to determine gradients present in the image frame. An offset mask can be determined from the image data and that offset mask can be used to adjust subsequently acquired image frames to reduce or remove gradients.

First claim

What is claimed is:

1. A method for adjusting an image using one or more images of locally flat scenes acquired with an imaging system comprising an imaging sensor with an array of photodetectors, the method comprising:
acquiring image data with the array of photodetectors, the image data comprising an array of pixel intensity values;
dividing at least a portion of the image data into a plurality of kernels;
determining local contrast values for individual kernels, a local contrast value of an individual kernel based at least in part on pixel values within the individual kernel;
determining that the image frame is a locally flat image frame if the local contrast values for at least a predetermined number of the individual kernels are less than or equal to a contrast threshold; and
in response to determining that the image frame is a locally flat image frame:
calculating an array of pixel adjustment values to generate an offset mask configured to adjust the pixel intensity values of the image data, wherein the pixel adjustment values are at least one of:
the value of each individual pixel in the locally flat image frame; or
a value representing a kernel of pixels from the locally flat image frame; and

adjusting pixel intensity values of subsequently acquired image data using the offset mask.

2. The method of claim 1 , wherein a pixel adjustment value in the array of pixel adjustment values of the offset mask corresponds to an average or median of the pixel intensity values within an individual kernel.

3. The method of claim 1 , wherein pixel adjustment values based on individual pixels or small kernels correspond to an offset mask representing a high spatial frequency non-uniformity correction and pixel adjustment values based on large kernels correspond to an offset mask representing a low spatial frequency correction including a correction for image gradients.

4. The method of claim 3 , wherein the large kernels are 3 by 3 pixels or larger.

5. The method of claim 3 , wherein both a low spatial frequency offset mask and a high spatial frequency offset mask are generated, and wherein both the low spatial frequency and high spatial frequency offset masks are applied to subsequently acquired image data at least one at a time or in combination.

6. The method of claim 5 , further comprising:
determining that a plurality of image frames over time are locally flat image frames; and
responsive to determining that a plurality of image frames over time are locally flat image frames, generating the offset mask using a statistical smoothing function applied to pixel adjustment values acquired from the plurality of locally flat image frames.

7. The method of claim 1 further comprising dividing individual kernels into a plurality of sub-kernels.

8. The method of claim 7 , wherein determining a local contrast value for an individual kernel comprises:
determining an average or median of pixel intensity values within each sub-kernel of the individual kernel; and
calculating a difference between a maximum sub-kernel average or median and a minimum sub-kernel average or median.

9. The method of claim 8 , wherein the average or median is adjusted using a bi-linear interpolation between adjacent kernels.

10. The method of claim 1 , wherein the offset mask is generated if the local contrast values for at least 90% of the plurality of kernels are less than or equal to a contrast threshold.

11. The method of claim 10 , wherein the offset mask is generated if the local contrast values for all of the plurality of kernels are less than or equal to a contrast threshold.

12. The method of claim 1 , wherein adjusting pixel intensity values using the offset mask comprises subtracting pixel adjustment values from corresponding pixel intensity values.

13. The method of claim 1 further comprising determining an offset mask offset comprising an average or median value of a kernel located near a center of the imaging array.

14. The method of claim 13 , wherein the offset mask includes the offset mask offset.

15. A thermal imaging system comprising:
an imaging array comprising an infrared focal plane array, the infrared focal plane array configured to generate signals corresponding to levels of infrared light incident on the infrared focal plane array;
a detector circuit comprising readout electronics that receive the generated signals and output image data comprising an array of pixel intensity values; and
a system controller configured to:
divide at least a portion of the image data into a plurality of kernels;
determine local contrast values for individual kernels, a local contrast value of an individual kernel based at least in part on pixel intensity values within the individual kernel;
determine that the image data corresponds to a locally flat scene if the local contrast values for at least a predetermined number of the individual kernels are less than or equal to a contrast threshold; and
in response to determining that the image data corresponds to a locally flat scene:
calculate an array of pixel adjustment values to generate an offset mask configured to adjust the pixel intensity values of the image data, the pixel adjustment values comprising at least one of: the value of each individual pixel in the image data, or a value representing a kernel of pixels from the image data; and
adjust pixel intensity values of subsequently acquired image data using the offset mask.

16. The thermal imaging system of claim 15 , wherein a pixel adjustment value in the array of pixel adjustment values of the offset mask corresponds to an average or median of the pixel intensity values within an individual kernel.

17. The thermal imaging system of claim 15 , wherein pixel adjustment values based on individual pixels or kernels smaller than 3 by 3 pixels correspond to an offset mask representing a high spatial frequency non-uniformity correction and pixel adjustment values based on kernels of at least 3 by 3 pixels correspond to an offset mask representing a low spatial frequency correction including a correction for image gradients.

18. The thermal imaging system of claim 17 , wherein the system controller is configured to generate both a low spatial frequency offset mask and a high spatial frequency offset mask, and to apply both the low spatial frequency and high spatial frequency offset masks to subsequently acquired image data at least one at a time or in combination.

19. The thermal imaging system of claim 15 , wherein the system controller is further configured to determine that a plurality of scenes over time are locally flat scenes and, responsive to determining that a plurality of scenes over time are locally flat scenes, generate the offset mask using a statistical smoothing function applied to pixel adjustment values acquired from the plurality of locally flat image frames.

20. The thermal imaging system of claim 15 , wherein the system controller is further configured to divide individual kernels into sub-kernels and to determine an average or median value of pixel intensities within each sub-kernel.

Assignees

• Seek Thermal Inc

Inventors

• Lee Jeffrey
• Parrish William J
• Engberg Andreas
• Williams Ross

Classifications

• H04N25/674 Primary

  based on the scene itself, e.g. defocusing · CPC title

• G06T2207/10048

  Infrared image · CPC title

• G06T2207/20021

  Dividing image into blocks, subimages or windows · CPC title

• G06T5/40

  using histogram techniques · CPC title

• H04N5/33

  Transforming infrared radiation (cameras or camera modules for generating image signals from infrared radiation H04N23/20; circuitry of SSIS for transforming infrared radiation into image signals H04N25/20) · CPC title