Recognition-based color corrections

Patent metadata

Field	Value
Publication number	US-10108876-B2
Application number	US-201615019976-A
Country	US
Kind code	B2
Filing date	Feb 10, 2016
Priority date	Feb 10, 2016
Publication date	Oct 23, 2018
Grant date	Oct 23, 2018

Abstract

A method comprising: performing a first object recognition round on an image to detect at least a first object; matching the first detected object to a first reference object, thereby recognizing the first object; determining a chromatic adaptation transform between the first recognized object and the first reference object; applying the chromatic adaptation transform to the image; performing a second object recognition round on the chromatically adapted image to detect a second object that is different than the first recognized object; and matching the second detected object with a second reference object, thereby recognizing the second object.

First claim

What is claimed is:

1. A method for computer-automated recognition of objects in digital imagery, comprising:
performing a first computer-automated object recognition round on a digital image to recognize at least a first object,
wherein recognizing the first object comprises detecting the first object and matching the first object with a first reference object, and
wherein performing the first computer-automated object recognition round comprises activating a texture-based object recognition engine on a grayscale version of the digital image;

determining a chromatic adaptation transform between the first recognized object and the first reference object;
applying the chromatic adaptation transform to the digital image; and
performing a second computer-automated object recognition round on the chromatically adapted digital image to recognize a second object that is different than the first recognized object,
wherein recognizing the second object comprises detecting the second object and matching the second object with a second reference object and
wherein performing the second computer-automated object recognition round comprises applying a color-based object recognition algorithm on the digital image.

2. The method of claim 1 , further comprising:
using the computer-automated recognition of the first and second objects to receive data stored in association with any of the first and second reference objects; and
rendering the associated data at an output interface.

3. The method of claim 1 , further comprising capturing the digital image.

4. A computer program product comprising a non-transitory computer-readable storage medium having program code embodied thereon, the program code executable by at least one hardware processor to:
perform a first computer-automated object recognition round on a digital image to recognize at least a first object,
wherein recognizing the first object comprises detecting the first object and matching the first object with a first reference object, and
wherein performing the first computer-automated object recognition round comprises activating a texture-based object recognition engine on a grayscale version of the digital image;

determine a chromatic adaptation transform between the first recognized object and the first reference object;
apply the chromatic adaptation transform to the digital image; and
perform a second computer-automated object recognition round on the chromatically adapted digital image to recognize a second object that is different than the first recognized object,
wherein recognizing the second object comprises detecting the second object and matching the second object with a second reference object, and
wherein performing the second computer-automated object recognition round comprises applying a color-based object recognition algorithm on the digital image.

5. The computer program product of claim 4 , wherein the program code is further executable to use the computer-automated recognition of the first and second objects to receive data stored in association with any of the first and second reference objects and render the associated data at an output interface.

6. The computer program product of claim 4 , wherein determining the chromatic adaptation transform comprises:
projectively transforming any of the first detected object and the first reference object to achieve a registration therebetween;
identifying one or more corresponding regions on each of the first recognized object and the first reference object having a stable color; and
computing a chromatic adaptation transform between the one or more corresponding regions of the first reference object and the first recognized object.

7. The computer program product of claim 6 , wherein the chromatic adaptation transform is computed using an iterative reweighted least squares solution.

8. The computer program product of claim 6 , wherein the program code is further executable to apply a low-pass filter to any of: the first recognized object, the first reference object, and any portion of the first recognized object and the first reference object, to remove outlying color values.

9. A method for computer-automated recognition of objects in digital imagery, comprising:
performing a first computer-automated object recognition round on a digital image to recognize at least a first object,
wherein recognizing the first object comprises detecting the first object and matching the first object with a first reference object;
determining a chromatic adaptation transform between the first recognized object and the first reference object,
wherein determining the chromatic adaptation transform comprises:
projectively transforming any of the first recognized object and the first reference object to achieve a registration therebetween,
identifying one or more corresponding regions on each of the first recognized object and the first reference object having a stable color, and
computing a chromatic adaptation transform between the one or more corresponding regions of the first reference object and the first recognized object;

applying the chromatic adaptation transform to the digital image; and
performing a second computer-automated object recognition round on the chromatically adapted digital image to recognize a second object that is different than the first recognized object,
wherein recognizing the second object comprises detecting the second object and matching the second object with a second reference object.

10. The method of claim 9 , wherein the chromatic adaptation transform is computed using an iterative reweighted least squares method.

11. The method of claim 9 , further comprising applying a low-pass filter to any of: the first recognized object, the first reference object, and any portion of the first recognized object and the first reference object, to remove outlying color values.

Assignees

• Shtok Joseph
• Tzadok Asaf
• IBM

Inventors

• Shtok Joseph
• Tzadok Asaf

Classifications

• H04N1/6033 Primary

  using test pattern analysis (H04N1/6055 takes precedence) · CPC title

• G06V10/56

  relating to colour · CPC title

• H04N23/88

  for colour balance, e.g. white-balance circuits or colour temperature control · CPC title

• G06V10/20

  Image preprocessing · CPC title

• G06K9/4652 Primary

  Physics · mapped topic