Apparatus and method for interband denoising and sharpening of images

Patent metadata

Field	Value
Publication number	US-12254601-B2
Application number	US-202217586435-A
Country	US
Kind code	B2
Filing date	Jan 27, 2022
Priority date	Jul 15, 2021
Publication date	Mar 18, 2025
Grant date	Mar 18, 2025

Abstract

A method includes obtaining a blended red-green-blue (RGB) image frame of a scene. The method also includes performing, using at least one processing device of an electronic device, an interband denoising operation to remove at least one of noise and one or more artifacts from the blended RGB image frame in order to produce a denoised RGB image frame. Performing the interband denoising operation includes performing filtering of red, green, and blue color channels of the blended RGB image frame to remove at least one of the noise and the one or more artifacts from the blended RGB image frame. The filtering of the red and blue color channels of the blended RGB image frame is based on image data of at least one of the green color channel and a white color channel of the blended RGB image frame.

First claim

What is claimed is:

1. A method comprising:
obtaining a blended red-green-blue (RGB) image frame of a scene; and
performing, using at least one processing device of an electronic device, an interband denoising operation to remove at least one of noise and one or more artifacts from the blended RGB image frame in order to produce a denoised RGB image frame;
wherein performing the interband denoising operation comprises performing filtering of red, green, and blue color channels of the blended RGB image frame to remove at least one of the noise and the one or more artifacts from the blended RGB image frame;
wherein the filtering of the green color channel of the blended RGB image frame produces a denoised green color channel; and
wherein the filtering of the red and blue color channels of the blended RGB image frame is based on image data of the denoised green color channel.

2. The method of claim 1 , wherein performing the filtering of the red, green, and blue color channels of the blended RGB image frame comprises:
filtering the green color channel of the blended RGB image frame using a first filter;
filtering the red color channel of the blended RGB image frame using a second filter; and
filtering the blue color channel of the blended RGB image frame using a third filter.

3. The method of claim 2 , wherein:
the first filter comprises a first guided filter configured to filter the green color channel of the blended RGB image frame while guided by one of: (i) a green-white channel based on the green color channel and a white color channel of the blended RGB image frame or (i) the green color channel of the blended RGB image frame;
the second filter comprises a second guided filter configured to filter the red color channel of the blended RGB image frame while guided by a combination of (i) the green-white channel or the green color channel of the blended RGB image frame, (ii) the red color channel of the blended RGB image frame, and (iii) the blue color channel of the blended RGB image frame; and
the third filter comprises a third guided filter configured to filter the blue color channel of the blended RGB image frame while guided by the combination of (i) the green-white channel or the green color channel of the blended RGB image frame, (ii) the red color channel of the blended RGB image frame, and (iii) the blue color channel of the blended RGB image frame.

4. The method of claim 2 , wherein:
the first filter comprises a first guided filter configured to filter the green color channel of the blended RGB image frame while guided by the green color channel of the blended RGB image frame, the first guided filter configured to generate the denoised green color channel;
the second filter comprises a second guided filter configured to filter the red color channel of the blended RGB image frame while guided by a combination of the red color channel of the blended RGB image frame and the denoised green color channel; and
the third filter comprises a third guided filter configured to filter the blue color channel of the blended RGB image frame while guided by a combination of the blue color channel of the blended RGB image frame and the denoised green color channel.

5. The method of claim 1 , wherein performing the filtering of the red, green, and blue color channels of the blended RGB image frame comprises:
filtering the green color channel of the blended RGB image frame using a green-channel denoising function;
filtering the red color channel of the blended RGB image frame using a first filter; and
filtering the blue color channel of the blended RGB image frame using a second filter.

6. The method of claim 5 , wherein:
the green-channel denoising function produces the denoised green color channel;
the first filter comprises a first guided filter configured to filter the red color channel of the blended RGB image frame while guided by a combination of the red color channel of the blended RGB image frame and the denoised green color channel; and
the second filter comprises a second guided filter configured to filter the blue color channel of the blended RGB image frame while guided by a combination of the blue color channel of the blended RGB image frame and the denoised green color channel.

7. The method of claim 1 , further comprising:
performing an interband sharpening operation to sharpen different color channels of the denoised RGB image frame in order to produce a sharpened RGB image frame.

8. The method of claim 7 , wherein performing the interband sharpening operation comprises:
identifying high-frequency contents of at least one of a green color channel and a white color channel of a reference image frame;
clipping the identified high-frequency contents based on one or more local maxima and one or more local minima in the different color channels of the denoised RGB image frame; and
combining the clipped high-frequency contents with the different color channels of the denoised RGB image frame in order to produce the sharpened RGB image frame.

9. The method of claim 1 , wherein the blended RGB image frame comprises an image frame generated using multiple input image frames that have been subjected to an intraband joint demosaic-warp operation.

10. The method of claim 1 , wherein the filtering of the green channel is based on image data of a white color channel.

11. The method of claim 1 , further comprising:
equalizing a brightness level of a white color channel compared to the green color channel to produce an equalized white color channel;
wherein the filtering of the green color channel is based on the equalized white color channel.

12. An electronic device comprising:
at least one processing device configured to:
obtain a blended red-green-blue (RGB) image frame of a scene;
perform an interband denoising operation to remove at least one of noise and one or more artifacts from the blended RGB image frame in order to produce a denoised RGB image frame; and
perform an interband sharpening operation to sharpen different color channels of the denoised RGB image frame in order to produce a sharpened RGB image frame;

wherein, to perform the interband denoising operation, the at least one processing device is configured to perform filtering of red, green, and blue color channels of the blended RGB image frame to remove at least one of the noise and the one or more artifacts from the blended RGB image frame;
wherein the at least one processing device is configured to perform the filtering of the red and blue color channels of the blended RGB image frame based on image data of at least one of the green color channel and a white color channel of the blended RGB image frame; and
wherein, to perform the interband sharpening operation, the at least one processing device is configured to:
identify high-frequency contents of at least one of a green color channel and a white color channel of a reference image frame;
clip the identified high-frequency contents based on one or more local maxima and one or more local minima in the different color channels of the denoised RGB image frame; and
combine the clipped high-frequency contents with the different color channels of the denoised RGB image frame in order to produce the sharpened RGB image frame.

13. The electronic device of claim 12 , wherein, to perform the filtering of the red, green, and blue color channels of the blended RGB image frame, the at least one processing device is configured to:
filter the green color channel of the blended RGB image frame using a first filter;
filter the red color channel of the blended RGB image frame using a second filter; and
f

Assignees

• Samsung Electronics Co Ltd

Inventors

• Le Long N
• Xu Yibo
• Glotzbach John W
• Tong Zhen
• Sheikh Hamid R

Classifications

• H04N25/683

  by defect estimation performed on the scene signal, e.g. real time or on the fly detection · CPC title

• G06T5/10

  using non-spatial domain filtering · CPC title

• H04N23/85

  for matrixing · CPC title

• H04N9/646

  for image enhancement, e.g. vertical detail restoration, cross-colour elimination, contour correction, chrominance trapping filters · CPC title

• G06T2207/10024

  Color image · CPC title