Image processing apparatus, image processing method, and recording medium

Patent metadata

Field	Value
Publication number	US-10007678-B2
Application number	US-201514942619-A
Country	US
Kind code	B2
Filing date	Nov 16, 2015
Priority date	Nov 21, 2014
Publication date	Jun 26, 2018
Grant date	Jun 26, 2018

Abstract

An image processing apparatus, an image processing method, and a recording medium realize identification of an object region and a region category with high precision, estimate a relationship between partial regions in an acquired image by using a database about relationships between the images, and determine a target region in the acquired image based on the relationship.

First claim

What is claimed is:

1. An image processing apparatus comprising:
at least one processor; and
a memory storing instructions that, when executed, cause the at least one processor to:
acquire an image;
divide the acquired image into a plurality of partial regions within an image frame of the acquired image;
extract one or more pieces of similar image data from each of the plurality of divided partial regions;
determine a target region, in the image frame of the acquired image, including partial regions having a relationship with each other indicative of a common object, based on information about a relationship of the acquired image the extracted image data; and
collate whether a plurality of the regions, in a plurality of the acquired images, including partial regions having a relationship with each other is related to each other based on the information used for determining the relative positional relationship between the plurality of the partial regions.

2. The image processing apparatus according to claim 1 , wherein a storage device is included in the image processing apparatus.

3. The image processing apparatus according to claim 1 , wherein the instructions further cause the at least one processor to:
output information about the region including partial regions having the relationship, with respect to the acquired image.

4. The image processing apparatus according to claim 1 ,
wherein the information about the relationship of a plurality of pieces of the image data includes information used for determining a correspondence relationship between the plurality of partial regions.

5. The image processing apparatus according to claim 4 ,
wherein the information about the relationship of the plurality of pieces of the image data includes information about graph data including a node that is one of a plurality of partial regions produced by dividing the plurality of pieces of the image data and an edge representing a correspondence relationship between the partial regions into which the image data is divided.

6. The image processing apparatus according to claim 5 ,
wherein the graph data includes an edge that is weighted based on similarity of the partial regions in the plurality of pieces of the image data.

7. The image processing apparatus according to claim 6 ,
wherein the weighted edge is acquired based on a teacher value about a category of an object in the plurality of pieces of the image data.

8. The image processing apparatus according to claim 5 ,
wherein the information about the relationship of the plurality of pieces of the image data is information about a subgroup acquired by clustering the plurality of pieces of the image data based on the graph data.

9. The image processing apparatus according to claim 8 , wherein the instructions further cause the at least one processor to:
identify a category of the region including partial regions having the relationship with each other based on the information about the subgroup.

10. The image processing apparatus according to claim 9 ,
wherein the identification unit further uses a trained identification dictionary to identify a category of the region including partial regions having the relationship with each other.

11. The image processing apparatus according to claim 5 ,
wherein the information about the relationship of the plurality of pieces of the image data includes information for determining an incorrect correspondence relationship between the plurality of partial regions.

12. The image processing apparatus according to claim 1 , wherein the information about the relationship of the plurality of pieces of the image data includes meta-information about the image in the plurality of pieces of the image data; and
wherein the image processing apparatus identifies a category of the region including partial regions having a relationship with each other, based on the meta-information about the image.

13. The image processing apparatus according to claim 12 ,
wherein the meta-information is an annotation label about an object or a portion of the object in the plurality of pieces of the image data.

14. The image processing apparatus according to claim 1 ,
wherein the information about the relationship of the plurality of pieces of the image data is information used for determining a relative positional relationship between the plurality of the partial regions.

15. An image processing apparatus comprising:
at least one processor; and
a memory storing instructions that, when executed, cause the at least one processor to:
acquire an image;
divide the acquired image into a plurality of partial regions within an image of the acquired image;
extract one or more pieces of similar image data from each of the plurality of divided partial regions;
determine a target region, in the image frame of the acquired image, including partial regions having a relationship with each other indicative of a common object, based on a teacher value about a category of an object relating to the extracted image data;
collate whether a plurality of the regions, in a plurality of the acquired images, including partial regions having a relationship with each other is related to each other based on the teacher value used in determining a relative positional relationship between the plurality of the partial regions.

16. The image processing apparatus according to claim 15 , wherein the instructions further cause the at least one processor to determine a category of the determined region.

17. An image processing method comprising:
acquiring an image;
dividing the acquired image into a plurality of partial regions within an image frame of the acquired image;
extracting one or more pieces of similar image data from each of the plurality of divided partial regions;
determining a target region, in the image frame of the acquired image, including partial regions having a relationship with each other indicative of a common object, based on the extracted image data and information about a relationship of the image data; and collating whether a plurality of the regions, in a plurality of the acquired images, including partial regions having a relationship with each other is related to each other based on the information used in determining a relative positional relationship between the plurality of partial regions.

18. A non-transitory computer-readable recording medium that stores a program, that when executed, causes a computer to function as an image processing apparatus by executing a method, the method comprising:
acquiring an image;
dividing the acquired image into a plurality of partial regions within an image frame of the acquired image;
extracting one or more pieces of similar image data from each of the plurality of divided partial regions from a storage; and
determining a target region, in the image frame of the acquired image, including partial regions having a relationship with each other indicative of a common object, based on information about a relationship of the acquired image the extracted image data and collating whether a plurality of the regions, in a plurality of the acquired images, including partial region having a relationship with each other is related to each other based on the information used in determining a relative positional relationship between the plurality of partial regions.

Assignees

• Canon Kk

Inventors

• Tate Shunta
• Mori Katsuhiko

Classifications

• G06F16/50

  of still image data · CPC title

• G06V20/00 Primary

  Scenes; Scene-specific elements (control of digital cameras H04N23/60) · CPC title

• G06V10/768

  using context analysis, e.g. recognition aided by known co-occurring patterns · CPC title

• G06V10/23 Primary

  based on positionally close patterns or neighbourhood relationships · CPC title

• G06F17/30244 Primary

  Physics · mapped topic