Image processing apparatus, image processing method, and program

• US 10,540,566 B2
• Filed: 03/26/2013
• Issued: 01/21/2020
• Est. Priority Date: 06/29/2012
• Status: Active Grant

• Current Assignee: NEC Corporation
• Original Assignee: NEC Corporation
• Inventors: Higa, Kyota
• Primary Examiner(s): Hu, Fred H
• Application Number:
  US14/411,587
• Publication Number:
  US 20150161468A1
• Time in Patent Office: 2,492 Days
• Field of Search: None
• CPC Class Codes: G06F 18/24137 Distances to cluster centroïds
• CPC Class Codes: G06F 18/2415 based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
• CPC Class Codes: G06V 10/464 using a plurality of salient features, e.g. bag-of-words [BoW] representations
• CPC Class Codes: G06V 10/757 Matching configurations of points or features
• CPC Class Codes: G06V 10/764 using classification, e.g. of video objects

Abstract

Provided is an image processing apparatus, an image processing method, and a program, which are capable of accurate collation even when an image contains a number of identical or similar subjects. The image processing apparatus of the present invention has: first feature amount generating means for generating, with respect to a plurality of feature points to be detected from a first image, a first local feature amount group including local feature amounts representing feature amounts of a plurality of local regions containing the respective feature points, and a first coordinate position information group including coordinate position information; region dividing means for clustering the feature points of the first image based on the first coordinate position information group; and collation means for collating, in units of clusters, the first local feature amount group with a second local feature amount group formed from local feature amounts of feature points detected from a second image.

Claims

1. An image processing apparatus, comprising:

a non-transitory storage device storing instructions; and

one or more processors configured by the instructions to;

generate, with respect to a plurality of feature points to be detected from a first image, a first local feature amount group including local feature amounts representing feature amounts of a plurality of local regions containing the respective feature points;

generate one or more first coordinate position information groups which include coordinate position information;

calculate a correspondence information group representing a correlation between a correlation between the feature points of the first image and feature points of a second image based on an inter-feature amount distance between the first local feature amount group and second local feature amount groups which is formed from local feature amounts of feature points detected from the second images;

cluster the feature points of the first image by using both the correspondence information groups and the first coordinate position information groups;

divide the first image into regions in accordance with the result of the clustering;

collate the first local feature amount groups for the respective regions of the first image with the second local feature amount group from the second image; and

identify different subjects within the first and second images based on the collated first and second local feature amount groups,

wherein the one or more processors are configured by the instructions to;

cluster the feature points of the first image based on a coordinate position of a predefined reference point, in the first image, of the second image, the coordinate position being estimated based on a relative coordinate position of each of the feature points of the second image and the reference point, the correspondence information groups, and the first coordinate position information groups; and

calculate a ratio between a distance between an two feature points of the first image and a distance between two feature points of the second image that correspond to the two feature points of the first image, by using the first coordinate position information groups, a second coordinate position information groups, which is coordinate position information on the feature points detected from the second image, and the correspondence information groups.

2. The image processing apparatus according to claim 1, wherein the one or more processors are configured by the instructions to cluster the feature points of the first image in accordance with an inter-feature point distance.

3. The image processing apparatus according to claim 1, wherein the one or more processors are configured by the instructions to cluster the feature points of the first image by using a degree of similarity between the local feature amounts of the first local feature amount groups and the first coordinate position information groups.

4. The image processing apparatus according to claim 1, wherein the one or more processors are configured by the instructions to:

calculate a rotation amount of a subject of the first image by using the first coordinate position information groups, the second coordinate information groups, and the correspondence information groups.

5. The image processing apparatus according to claim 4, wherein the one or more processors are configured by the instructions to cluster the feature points of the first image by using the first coordinate position information groups and either one of the ratio and the rotation amount.

6. The image processing apparatus according to claim 4, wherein the one or more processors are configured by the instructions to cluster the feature points of the first image by using the coordinate position, in the first image, of the reference point of the second image, the coordinate position being estimated using the rotation amount, the ratio, the relative coordinate position, and the first coordinate position information groups.

7. An image processing method, comprising the steps of:

generating, with respect to a plurality of feature points to be detected from a first image, a first local feature amount group including local feature amounts representing feature amounts of a plurality of local regions containing the respective feature points;

generating one or more first coordinate position information groups which include coordinate position information;

calculating a correspondence information group representing a correlation between a correlation between the feature points of the first image and feature points of a second image based on an inter-feature amount distance between the first local feature amount group and second local feature amount groups which is formed from local feature amounts of feature points detected from the second images;

clustering the feature points of the first image by using both the correspondence information groups and the first coordinate position information groups;

dividing the first image into regions in accordance with the result of the clustering;

collating the first local feature amount groups for the respective regions of the first image with the second local feature amount group from the second image;

identifying different subjects within the first and second images based on the collated first and second local feature amount groups,

cluster the feature points of the first image based on a coordinate position of a predefined reference point, in the first image, of the second image, the coordinate position being estimated based on a relative coordinate position of each of the feature points of the second image and the reference point, the correspondence information groups, and the first coordinate position information groups; and

calculate a ratio between a distance between any two feature points of the first image and a distance between two feature points of the second image that correspond to the two feature points of the first image, by using the first coordinate position information groups, a second coordinate position information groups, which is coordinate position information on the feature points detected from the second image, and the correspondence information groups.

8. A non-transitory computer readable storage medium storing a program for causing an image processing apparatus to execute the steps of:

generating, with respect to a plurality of feature points to be detected from a first image, a first local feature amount group including local feature amounts representing feature amounts of a plurality of local regions containing the respective feature points;

generating one or more first coordinate position information groups which include coordinate position information;

calculating a correspondence information group representing a correlation between a correlation between the feature points of the first image and feature points of a second image based on an inter-feature amount distance between the first local feature amount group and second local feature amount groups which is formed from local feature amounts of feature points detected from the second images;

clustering the feature points of the first image by using both the correspondence information groups and the first coordinate position information groups;

dividing the first image into regions in accordance with the result of the clustering;

collating, the first local feature amount groups for the respective regions of the first image with the second local feature amount group from the second image; and

identifying different subjects within the first and second images based on the collated first and second local feature amount groups,

cluster the feature points of the first image based on a coordinate position of a predefined reference point, in the first image, of the second image, the coordinate position being estimated based on a relative coordinate position of each of the feature points of the second image and the reference point, the correspondence information groups, and the first coordinate position information groups; and

calculate a ratio between a distance between an two feature points of the first image and a distance between two feature points of the second image that correspond to the two feature points of the first image, by using the first coordinate position information groups, a second coordinate position information groups, which is coordinate position information on the feature points detected from the second image, and the correspondence information groups.