DEPTH INFORMATION EXTRACTING METHOD BASED ON MACHINE LEARNING AND APPARATUS THEREOF

• US 20170069094A1
• Filed: 09/02/2016
• Published: 03/09/2017
• Est. Priority Date: 09/04/2015
• Status: Active Grant

• Current Assignee: Hyundai Motor Company (Hyundai Motor Group), Kia Corp.
• Original Assignee: Electronics and Telecommunications Research Institute
• Inventors: KIM, Hye-Jin
• Granted Patent: US 10,043,285 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 18/2413 based on distances to training or reference patterns
• CPC Class Codes: G06T 2207/10024 Color image
• CPC Class Codes: G06T 2207/20081 Training; Learning
• CPC Class Codes: G06T 2207/20084 Artificial neural networks [ANN]
• CPC Class Codes: G06T 7/529 from texture
• CPC Class Codes: G06V 10/764 using classification, e.g. of video objects
• CPC Class Codes: G06V 10/82 using neural networks
• CPC Class Codes: G06V 20/647 by matching two-dimensional images to three-dimensional objects

Abstract

The disclosure relates to a method and an apparatus for extracting depth information from an image. A method for extracting depth information based on machine learning according to an exemplary embodiment of the present disclosure includes generating a depth information model corresponding to at least one learning image by performing machine learning using the at least one learning image and a plurality of depth information corresponding to the at least one learning image; and extracting depth information of a target image by applying the depth information model into the target image. Embodiments of the disclosure may allow extracting precise depth information from a target image.

Claims

1. A method for extracting depth information based on machine learning comprising:

generating a depth information model corresponding to at least one learning image by performing machine learning using the at least one learning image and a plurality of depth information corresponding to the at least one learning image; and

extracting depth information of a target image by applying the depth information model into the target image.

2. The method of claim 1, wherein the generating of a depth information model comprises:

generating an N^th depth information submodel by performing machine learning using the at least one learning image and N^th depth information corresponding to the at least one learning image; and

generating the depth information model by performing machine learning using the generated N depth information submodels.

3. The method of claim 1, wherein the extracting of depth information comprises:

loading a depth information model corresponding to the target image and applying the loaded depth information model into the target image to extract depth information for the target image.

4. The method of claim 1, wherein the plurality of depth information is obtained through methods for extracting depth information which are different from one another.

5. The method of claim 1, further comprising generating a learning image having improved depth information by performing machine learning using the depth information model and the at least one learning image.

6. The method of claim 1, wherein the depth information model represents a relationship between the learning image and the plurality of depth information.

7. An apparatus for extracting depth information based on machine learning, comprising:

a depth information model learning unit configured to generate a depth information model corresponding to at least one learning image by performing machine learning using the at least one learning image and a plurality of depth information corresponding to the at least one learning image; and

a depth information sensing unit configured to extract depth information of a target image by applying the depth information model into the target image.

8. The apparatus of claim 7, wherein the depth information model learning unit generates an N^th depth information submodel by performing machine learning using the at least one learning image and the N^th depth information corresponding to the at least one learning image, and generates the depth information model by performing machine learning using the generated N depth information submodels.

9. The apparatus of claim 7, wherein the depth information sensing unit loads a depth information model corresponding to the target image and applies the loaded depth information model into the target image to extract depth information for the target image.

10. The apparatus of claim 7, further comprising a depth information obtaining unit configured to obtain the plurality of depth information.

11. The apparatus of claim 7, wherein the plurality of depth information is obtained through methods for extracting depth information which are different from one another.

12. The apparatus of claim 7, wherein the depth information model learning unit is configured to generate a learning image having improved depth information by performing machine learning using the depth information model and the at least one learning image.

13. The apparatus of claim 7, wherein the depth information model represents a relationship between the learning image and the plurality of depth information.

14. An apparatus for extracting depth information based on machine learning comprising:

a depth information model learning unit configured to generate a depth information model corresponding to a plurality of learning images by performing machine learning using the plurality of learning images and at least one depth information corresponding to the plurality of learning images; and

a depth information sensing unit configured to extract depth information of a target image by applying the depth information model into the target image.

15. The apparatus of claim 14, wherein the depth information model learning unit generates an N^th depth information submodel by performing machine learning using the at least one depth information and the N^th learning image corresponding to the at least one depth information, and generates the depth information model by performing machine learning using the generated N depth information submodels.

16. The apparatus of claim 14, wherein the depth information sensing unit loads a depth information model corresponding to the target image and applies the loaded depth information model into the target image to extract depth information for the target image.

17. The apparatus of claim 14, further comprising a depth information obtaining unit configured to obtain the plurality of depth information.

18. The apparatus of claim 14, wherein the plurality of depth information is obtained through methods for extracting depth information which are different from one another.

19. The apparatus of claim 14, wherein the depth information model learning unit is configured to generate a learning image having improved depth information by performing machine learning using the depth information model and the plurality of learning images.

20. The apparatus of claim 14, wherein the depth information model represents relationship between the learning image and the plurality of depth information.