Image retrieving device, image retrieving method, and recording medium

Patent metadata

Field	Value
Publication number	US-10360471-B2
Application number	US-201715398834-A
Country	US
Kind code	B2
Filing date	Jan 5, 2017
Priority date	Jan 27, 2016
Publication date	Jul 23, 2019
Grant date	Jul 23, 2019

Abstract

An image retrieving device capable of retrieving an image using a local feature quantity, the image retrieving device includes a memory, and a processor in communication with the memory, the processor being configured to perform operations including receiving a designation of a retrieval condition corresponding to a retrieval source image, determining whether to use a feature quantity of a local image having scale invariance on the basis of the retrieval condition, and performing control to calculate the feature quantity of the local image having scale invariance if it is determined to use the feature quantity of the local image having scale invariance, and performing control not to calculate the feature quantity of the local image having scale invariance if it is determined not to use the feature quantity of the local image having scale invariance.

First claim

What is claimed is:

1. An image retrieving device capable of retrieving an image using a local feature quantity, the image retrieving device comprising:
(a) a memory; and
(b) a processor in communication with the memory, the processor being configured to perform operations comprising:
(i) receiving a designation of a retrieval condition corresponding to a retrieval source image;
(ii) determining whether or not to use a feature quantity of a local image having scale invariance on the basis of an estimated retrieval time that is estimated based on the retrieval condition; and
(iii) performing control to calculate the feature quantity of the local image having scale invariance if it is determined to use the feature quantity of the local image having scale invariance, and performing control not to calculate the feature quantity of the local image having scale invariance if it is determined not to use the feature quantity of the local image having scale invariance.

2. The image retrieving device according to claim 1 , wherein a designation of a scale correspondence range is received as the designation of the retrieval condition.

3. The image retrieving device according to claim 2 , wherein, if it is determined not to use the feature quantity of the local image having scale invariance, a parasitic image is generated based on the scale correspondence range from the retrieval source image.

4. The image retrieving device according to claim 1 , wherein the processor further performs an operation comprising (iv) detecting feature points from an image and calculating the feature quantity of the local image using pixels neighboring the feature points.

5. The image retrieving device according to claim 1 , wherein the estimated retrieval time is estimated by calculating a retrieval time on the basis of the retrieval condition.

6. The image retrieving device according to claim 5 , wherein the processor further performs operations comprising:
(iv) calculating a first estimated retrieval time in a case when the feature quantity of the local image having scale invariance is used; and
(v) calculating a second estimated retrieval time in a case when the feature quantity of the local image having scale invariance is not used,
wherein it is determined to use the feature quantity of the local image having scale invariance in a case when the first estimated retrieval time is shorter than the second estimated retrieval time, and
wherein it is determined not to use the feature quantity of the local image having scale invariance in a case when the first estimated retrieval time is not shorter than the second estimated retrieval time.

7. The image retrieving device according to claim 5 , wherein the estimated retrieval time is calculated on the basis of at least one of the scale correspondence range corresponding to the retrieval source image, the number of retrieval source images, the number of parasitic images generated based on the scale correspondence range from the retrieval source images, and an average number of local feature points that is an average of the numbers of feature points in the retrieval target image.

8. The image retrieving device according to claim 7 , wherein the processor further performs operations comprising:
(iv) detecting feature points from an image and calculating local feature quantities of the feature points using pixels neighboring the feature points;
(v) storing the numbers of feature points of the retrieval target image;
(vi) calculating the average number of local feature points of the retrieval target image from the stored numbers of feature points of the retrieval target image; and
(vii) storing the average number of local feature points of the retrieval target image.

9. The image retrieving device according to claim 8 , wherein, if an absolute value of a difference between the latest average number of local feature points of the retrieval target image and the number of local feature points of the retrieval target image when image retrieval has previously been performed is greater than a threshold value, the latest average number of local feature points of the retrieval target image is used as the average number of local feature points of the retrieval target image for calculating the estimated retrieval time and the estimated retrieval time is calculated.

10. An image retrieving method comprising:
receiving a designation of a retrieval condition corresponding to a retrieval source image;
determining whether or not to use a feature quantity of a local image having scale invariance on the basis of an estimated retrieval time that is estimated based on the retrieval condition; and
performing control to calculate the feature quantity of the local image having scale invariance if it is determined to use the feature quantity of the local image having scale invariance, and performing control not to calculate the feature quantity of the local image having scale invariance if it is determined not to use the feature quantity of the local image having scale invariance.

11. A non-transitory computer-readable recording medium having stored a program for causing a computer to execute a method, the method comprising:
receiving a designation of a retrieval condition corresponding to a retrieval source image;
determining whether or not to use a feature quantity of a local image having scale invariance on the basis of an estimated retrieval time that is estimated based on the retrieval condition; and
performing control to calculate the feature quantity of the local image having scale invariance of it is determined to use the feature quantity of the local image having scale invariance, and performing control not to calculate the feature quantity of the local image having scale invariance if it is determined not to use the feature quantity of the local image having scale invariance.

Assignees

• Canon Kk

Inventors

• Arakawa Junya
• Otake Ritsuko

Classifications

• G06V10/945

  User interactive design; Environments; Toolboxes · CPC title

• G06F18/40

  Software arrangements specially adapted for pattern recognition, e.g. user interfaces or toolboxes therefor · CPC title

• G06V10/462 Primary

  Salient features, e.g. scale invariant feature transforms [SIFT] · CPC title

• G06V10/52

  Scale-space analysis, e.g. wavelet analysis (multi-scale boundary representations G06V10/42) · CPC title

• G06V10/758

  Involving statistics of pixels or of feature values, e.g. histogram matching · CPC title