Methods and systems for image processing

Patent metadata

Field	Value
Publication number	US-12056843-B2
Application number	US-202117304945-A
Country	US
Kind code	B2
Filing date	Jun 29, 2021
Priority date	Dec 29, 2018
Publication date	Aug 6, 2024
Grant date	Aug 6, 2024

Abstract

Methods and systems are provided for processing an image. The method may include selecting one or more tiles in a first image. The method may also include performing, for each of the one or more tiles, an image processing operation on an image processing region within the tile based on a second image for K iterations, decreasing the size of the image processing region with each iteration, wherein K is an integer above 2.

First claim

We claim:

1. A system for processing an image, the system comprising:
a storage medium, including a set of instructions; and
at least one processor in communication with the storage medium, when executing the set of instructions, the at least one processor is to:
select one or more tiles in a first image;
perform, for each of the one or more tiles, an image processing operation on an image processing region within the tile based on a second image for K iterations, decreasing the size of the image processing region with each iteration, wherein K is an integer above 2, and wherein
in each iteration of the K iterations, the image processing operation modifies pixel values of one or more pixels within the image processing region, the one or more modified pixels forming a processed region having a size smaller than the size of the corresponding image processing region; and
the at least one processor is further to:
designate the processed region of a current iteration as the image processing region of a next iteration.

2. The system of claim 1 , wherein to perform the image processing operation on the image processing region, in each iteration of the K iterations, the at least one processor is to:
set one or more window regions by running a sliding window on the image processing region, wherein each of the one or more window regions corresponds to a first block of pixels of the first image, the first block of pixels including a first pixel of interest; and
determine an output value for each of the one or more window regions to modify the first pixel of interest.

3. The system of claim 2 , wherein to determine the output value for each of the one or more window regions, the at least one processor is to:
obtain, from the second image, a second block of pixels corresponding to the first block of pixels, the second block of pixels including a second pixel of interest corresponding to the first pixel of interest;
compute a difference indicator indicating a difference between the first block of pixels and the second block of pixels; and
compute the output value based on at least the difference indicator, a pixel value of the first pixel of interest, and a pixel value of the second pixel of interest.

4. The system of claim 3 , wherein:
the first image and the second image are formatted according to a filter color array.

5. The system of claim 4 , further comprising:
an image sensor covered by the color filter array, wherein the first image and the second image are generated by the image sensor.

6. The system of claim 4 , wherein to compute the difference indicator, the at least one processor is to:
compute a first mean of pixel values of pixels having a predetermined color in the first block of pixels;
determine a second mean of pixel values of pixels having the predetermined color in the second block of pixels; and
determine the difference indicator at least based on a difference between the first mean and the second mean.

7. The system of claim 3 , wherein to compute the output value, the at least one processor is to:
obtain a first high-frequency component of the first pixel of interest and a first low-frequency component of the first pixel of interest;
obtain a second high-frequency component of the second pixel of interest;
compute an updated high-frequency component of the first pixel of interest at least based on the first difference indicator, the first high-frequency component, and the second high-frequency component; and
obtain the output value based on the updated high-frequency component of the first pixel of interest and the first low-frequency component.

8. The system of claim 7 , wherein to compute the updated high-frequency component of the first pixel of interest, the at least one processor is to:
obtain, based at least on the first difference indicator, weights associated with the first pixel of interest and the second pixel of interest; and
compute a weighted sum or weighted mean of the original high-frequency component of the first pixel of interest and the high-frequency component of the second pixel of interest as the updated high-frequency component.

9. The system of claim 8 , wherein in at least one iteration of the K iterations, for each of the one or more window region, the at least one processor is further to:
determine a brightness indicator indicating a brightness of the first block of pixels, wherein the first weight and the second weight are determined based further on the brightness indicator.

10. The system of claim 8 , wherein:
the second image is obtained by processing a third image;
in at least one iteration of the K iterations, for each of the one or more window regions, the first weight and the second weight are determined based further on a reference parameter associated with the second pixel of interest; and
the at least one processor is further to, in the Kth iteration of the K iterations, for each of the one or more window regions, store the first difference indicator as the reference parameter associated with the first pixel of interest.

11. A method for processing an image, implemented on an image processing device, the method comprising:
selecting, by the image processing device, one or more tiles in a first image;
performing, by the image processing device, for each of the one or more tiles, an image processing operation on an image processing region within the tile based on a second image for K iterations, decreasing the size of the image processing region with each iteration, wherein K is an integer above 2, and wherein
in each iteration of the K iterations, the image processing operation modifies pixel values of one or more pixels within the image processing region, the one or more modified pixels forming a processed region having a size smaller than the size of the corresponding image processing region; and
the method further comprises:
designating the processed region of a current iteration as the image processing region of a next iteration.

12. The method of claim 11 , wherein the performing an image processing operation on an image processing region includes, in each iteration of the K iterations:
setting one or more window regions by running a sliding window on the image processing region, wherein each of the one or more window regions corresponds to a first block of pixels of the first image, the first block of pixels including a first pixel of interest; and
determining an output value for each of the one or more window regions to modify the first pixel of interest.

13. The method of claim 12 , wherein the determining an output value for each of the one or more window regions comprises:
obtaining, from the second image, a second block of pixels corresponding to the first block of pixels, the second block of pixels including a second pixel of interest corresponding to the first pixel of interest;
computing a difference indicator indicating a difference between the first block of pixels and the second block of pixels; and
computing the output value based on at least the difference indicator, a pixel value of the first pixel of interest, and a pixel value of the second pixel of interest.

14. The method of claim 13 , wherein the computing the output value includes:
obtaining a first high-frequency component of the first pixel of interest and a first low-frequency component of the first pixel of interest;
obtaining a second high-frequency component of the second pixel of interest;
computing an updated high-frequency component of the first pixel of interest at least based on the first difference indicator, the first high-frequency component,

Assignees

• Zhejiang Dahua Technology Co

Inventors

• Ling Chen
• Ma Xin
• Fang Wei

Classifications

• G06T3/4084

  in the transform domain, e.g. fast Fourier transform [FFT] domain scaling · CPC title

• G06T7/90

  Determination of colour characteristics · CPC title

• G06T3/4015 Primary

  Image demosaicing, e.g. colour filter arrays [CFA] or Bayer patterns · CPC title

• G06T2207/20012

  Locally adaptive · CPC title

• G06T2207/20016

  Hierarchical, coarse-to-fine, multiscale or multiresolution image processing; Pyramid transform · CPC title