Local feature extraction apparatus, control method therefor, and computer-readable medium

• US 8,842,917 B2
• Filed: 09/27/2010
• Issued: 09/23/2014
• Est. Priority Date: 11/05/2009
• Status: Active Grant

• Current Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Original Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Inventors: Magai, Koichi, Shiiyama, Hirotaka, Ikeda, Kazuyo
• Primary Examiner(s): Koziol, Stephen R
• Assistant Examiner(s): CONNER, SEAN M
• Application Number:
  US12/891,486
• Publication Number:
  US 20110103697A1
• Time in Patent Office: 1,457 Days
• Field of Search: None
• US Class Current: 382/195
• CPC Class Codes: G06V 10/462 Salient features, e.g. scale invariant feature transforms [SIFT]

Abstract

An information processing apparatus includes: a calculation unit adapted to analyze an image and calculate an intermediate value; a setting unit adapted to set a feature extraction region in the image, using the intermediate value; and an extraction unit adapted to extract a local feature of the feature extraction region, reusing the intermediate value used by the setting unit.

Claims

1. An information processing apparatus comprising:

an obtaining unit configured to analyze an image and obtain an intermediate value;

a setting unit configured to set a feature extraction region in the image, using the intermediate value, wherein an area of the feature extraction region gradually expands after processing for setting the feature extraction region has started and until the processing has been completed;

an extraction unit configured to extract a feature point from the feature extraction region, by reusing the intermediate value used by the setting unit; and

an acquisition unit configured to acquire a local image feature amount of the feature point, by reusing the intermediate value used by the setting unit,

wherein, after the processing for setting the feature extraction region has started, the setting unit measures the area of the feature extraction region and, when the area is greater than or equal to a predetermined value, stops the processing and sets the feature extraction region to the entire image.

2. The information processing apparatus according to claim 1, wherein the obtaining unit obtains a luminance gradient value for each pixel location in the image.

3. The information processing apparatus according to claim 2, wherein the setting unit sets the feature extraction region to a region where the luminance gradient value is greater than or equal to a predetermined threshold.

4. The information processing apparatus according to claim 1, wherein the obtaining unit performs frequency conversion on the image and obtains a peak frequency coefficient in an image obtained by the frequency conversion as the intermediate value.

5. The information processing apparatus according to claim 4, wherein the setting unit sets the feature extraction region to a region where the peak frequency coefficient is greater than or equal to a predetermined threshold.

6. The information processing apparatus according to claim 1, wherein the setting unit divides the image into a plurality of partial regions, selects the partial regions in random order, and sets the feature extraction region for each of the partial regions in the order of selection.

7. The information processing apparatus according to claim 1, further comprising:

a holding unit configured to hold each local image feature amount acquired by the acquisition unit in association with the image.

8. The information processing apparatus according to claim 1, wherein

the image is an input image,

the information processing apparatus further comprising:

a retrieval unit configured to retrieve an image similar to the input image by comparing a local image feature amount of the input image with a local image feature amount of each image held in a holding unit.

10. A non-transitory computer-readable storage medium having stored thereon a program for causing a computer to serve as each unit of the information processing apparatus according to claim 1.

9. A control method for an information processing apparatus, comprising:

an obtaining step, performed by an obtaining unit, of analyzing an image and obtaining an intermediate value;

a setting step, performed by a setting unit, of setting a feature extraction region in the image, using the intermediate value, wherein an area of the feature extraction region gradually expands after processing for setting the feature extraction region has started and until the processing has been completed; and

an extraction step, performed by an extraction unit, of extracting a feature point from the feature extraction region, by reusing the intermediate value used in the setting step;

an acquisition step, performed by an acquisition unit, of acquiring a local image feature amount of the feature point, by reusing the intermediate value used in the setting step,

wherein, after the processing for setting the feature extraction region has started, the setting step measures the area of the extraction region and, when the area is greater than or equal to a predetermined value, stops the processing and sets the feature extraction region to the entire image.