Method and apparatus for processing image

Patent metadata

Field	Value
Publication number	US-12164601-B2
Application number	US-202318460922-A
Country	US
Kind code	B2
Filing date	Sep 5, 2023
Priority date	Oct 28, 2021
Publication date	Dec 10, 2024
Grant date	Dec 10, 2024

Abstract

A method for processing an image, which is performed by an image processing apparatus, is provided. The method includes acquiring a first image including an object and a second image including an object identical to the object in the first image under the same condition, acquiring three-dimensional direction information of a specific part of the object in the first image, and providing a three-dimensionally processed image by three-dimensionally rotating the object in the second image by an angle that corresponds to the acquired three-dimensional direction information of the specific part of the object in the first image.

First claim

What is claimed is:

1. A method for processing an image performed by an image processing apparatus, the method comprising:
acquiring a first image including an object and a second image including the object in the first image under the same condition;
acquiring three-dimensional direction information of a specific part of the object in the first image; and
providing a three-dimensionally processed image by three-dimensionally rotating the object in the second image by an angle that corresponds to the acquired three-dimensional direction information of the specific part of the object in the first image,
wherein, in the providing the three-dimensionally processed image, a partial image including the specific part of the object in the second image is acquired, and then, a three-dimensionally processed partial image is provided by three-dimensionally rotating the partial image in reverse direction of the acquired three-dimensional direction information of the specific part of the object in the first image by the angle that corresponds to the acquired three-dimensional direction information of the specific part of the object in the first image, and
wherein an absolute value of the angle and that of an angle in the acquired three-dimensional direction information are substantially identical to each other.

2. The method of claim 1 , wherein the first image is a 2-dimensional image, and the second image is a 3-dimensional image.

3. The method of claim 2 , wherein the 2-dimensional image is a 2-dimensional infrared ray (IR) image or a 2-dimensional color image.

4. The method of claim 1 , wherein the object is a human body, and the specific part is a face of the human body.

5. The method of claim 1 , wherein, in the acquiring the three-dimensional direction information, the three-dimensional direction information is acquired by using an artificial intelligence neural network model which is trained with input data and label data, and
the input data includes a plurality of training 2-dimensional images, and the label data includes three-dimensional direction information of a specific part of an object included in each of the plurality of training 2-dimensional images.

6. The method of claim 1 , further comprises providing a comparison result between the rotation processed partial image and reference data.

7. The method of claim 6 , wherein the object is a human body, and the specific part is a face of the human body, and
wherein, in the providing the comparison result, the comparison result is information that indicates whether the rotation processed partial image and the reference data show the same face.

8. A method for processing an image performed by an image processing apparatus, the method comprising:
acquiring a first image including an object and a second image including an object identical to the object in the first image under the same condition;
acquiring direction information of a specific part of the object in the first image;
determining a partial image including the specific part of the object in the second image; and
rotating the partial image including the specific part of the object in the second image based on the acquired direction information of the specific part of the object in the first image to provide a rotation processed partial image,
wherein the direction information of the specific part of the object in the first image includes information for yaw direction, information for roll direction, and information for pitch direction of the specific part of the object,
wherein, in the acquiring the direction information, the direction information is acquired by using an artificial intelligence neural network model which is trained with input data and label data,
wherein the input data includes a plurality of training 2-dimensional images, and the label data includes a 3-dimensional object corresponding to an object included in each of the plurality of training 2-dimensional images and direction information of a specific part of the 3-dimensional object, and
wherein the rotating the partial image includes rotating the partial image including the specific part of the object in the second image in reverse direction of the acquired direction information of the specific part of the object in the first image by an angle that corresponds to the acquired direction information of the specific part of the object in the first image, and
wherein an absolute value of the angle and that of an angle in the acquired direction information of the specific part of the object in the first image are substantially identical to each other.

9. The method of claim 8 , wherein the first image is a 2-dimensional image, and the second image is a 3-dimensional image.

10. The method of claim 9 , wherein the 2-dimensional image is a 2-dimensional infrared ray (IR) image or a 2-dimensional color image.

11. The method of claim 8 , wherein the object is a human body, and the specific part is a face of the human body.

12. The method of claim 8 , further comprises providing a comparison result between the rotation processed partial image and reference data.

13. A non-transitory computer-readable storage medium including computer-executable instructions which cause, when executed by a processor, the processor to perform a method for processing an image performed by an image processing apparatus, the method comprising:
acquiring a first image including an object and a second image including an object identical to the object in the first image under the same condition;
acquiring direction information of a specific part of the object in the first image;
determining a partial image including the specific part of the object in the second image; and
rotating the partial image including the specific part of the object in the second image based on the acquired direction information of the specific part of the object in the first image to provide a rotation processed partial image,
wherein the direction information of the specific part of the object in the first image includes information for yaw direction, information for roll direction, and information for pitch direction of the specific part of the object,
wherein, in the acquiring the direction information, the direction information is acquired by using an artificial intelligence neural network model which is trained with input data and label data,
wherein the input data includes a plurality of training 2-dimensional images, and the label data includes a 3-dimensional object corresponding to an object included in each of the plurality of training 2-dimensional images and direction information of a specific part of the 3-dimensional object, and
wherein the rotating the partial image includes rotating the partial image including the specific part of the object in the second image in reverse direction of the acquired direction information of the specific part of the object in the first image by an angle that corresponds to the acquired direction information of the specific part of the object in the first image, and
wherein an absolute value of the angle and that of an angle in the acquired direction information of the specific part of the object in the first image are substantially identical to each other.

14. The method of claim 13 , wherein the first image is a 2-dimensional image, and the second image is a 3-dimensional image.

15. The method of claim 14 , wherein the 2-dimensional image is a 2-dimensional infrared ray (IR) image or a 2-dimensional color image.

16. The method of claim 13 , wherein the object is a human body, and the specific part is a face

Assignees

• Suprema Inc

Inventors

• Lee Hyogi
• Lee Kideok
• Song Bong Seop

Classifications

• G06N3/09

  Supervised learning · CPC title

• G06F18/2148

  characterised by the process organisation or structure, e.g. boosting cascade · CPC title

• G06V40/172 Primary

  Classification, e.g. identification · CPC title

• G06T2207/20081

  Training; Learning · CPC title

• G06T2207/20084

  Artificial neural networks [ANN] · CPC title