Information processing apparatus, method, and program

• US 8,103,115 B2
• Filed: 02/22/2008
• Issued: 01/24/2012
• Est. Priority Date: 02/26/2007
• Status: Expired due to Fees

• Current Assignee: Sony Corporation (Sony Group Corp.)
• Original Assignee: Sony Corporation (Sony Group Corp.)
• Inventors: Iwai, Yoshiaki, Yoshigahara, Takayuki, Nakamura, Akira
• Primary Examiner(s): Bella, Matthew
• Assistant Examiner(s): Rosario, Dennis
• Application Number:
  US12/070,956
• Publication Number:
  US 20080205764A1
• Time in Patent Office: 1,432 Days
• Field of Search: 382/171, 382/195, 382/199, 382/201, 382/203, 382/205, 382/246
• US Class Current: 382/246
• CPC Class Codes: G06V 10/421 by analysing segments intersecting the pattern
• CPC Class Codes: G06V 10/46 Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features colour feature extraction G06V10/56

Abstract

An information processing apparatus that compares an input image with a model image to identify the subject of the input image with the subject of the model image. The apparatus includes feature value extraction means for setting feature points, each of which is on an edge of the model image and provided to extract a model image feature value, which is the feature value of the model image, and extracting the model image feature value from each of a plurality of feature value extraction areas in the neighborhood of each of the feature points, and matching means for checking if an input image feature value, which is the feature value of the input image, at the point that is on an edge of the input image and corresponds to the feature point matches any of the plurality of model image feature values at the feature point.

Claims

1. An information processing apparatus that compares an input image with a model image to identify the subject of the input image with the subject of the model image and that includes a processor, the apparatus comprising:

feature value extraction means for setting feature points, each of which is on an edge of the model image and provided to extract a model image feature value, which is the feature value of the model image, and extracting the model image feature value from each of a plurality of feature value extraction areas in the neighborhood of each of the feature points; and

matching means for determining whether an input image feature value, which is the feature value of the input image, at the point that is on an edge of the input image and corresponds to a feature point matches any of the plurality of model image feature values at the feature point,

wherein the feature value extracting means selects a plurality of edge points as feature points based on an entropy value indicating how the plurality of edge points are scattered.

2. The information processing apparatus according to claim 1,

wherein the feature value extraction means extracts multiple types of the model image feature value from each of the plurality of feature value extraction areas, and

the matching means carries out the matching operation for each of the multiple types.

3. The information processing apparatus according to claim 1,

wherein the feature value extraction means extracts the model image feature value from each of the plurality of feature value extraction areas, the model image feature value having successive values for each pixel in the neighborhood of an edge of the model image.

4. The information processing apparatus according to claim 1,

wherein the feature value extraction means extracts an edge intensity from each of the plurality of feature value extraction areas, the edge intensity obtained in the neighborhood of an edge of the model image.

5. The information processing apparatus according to claim 1,

wherein the feature value extraction means extracts the model image feature value, which is the feature value of the model image, from each of the plurality of feature value extraction areas in the neighborhood of the feature point defined as the point where a reference circle determined in such a way that a relatively large number of edges of the model image are contained therein intersects an edge of the model image.

6. The information processing apparatus according to claim 1,

wherein the feature value extraction means extracts the feature value from each of small areas obtained by radially segmenting each of the feature value extraction areas that is formed of a plurality of concentric circles having different radii, each of the segmented areas having a predetermined angle, and surrounded by the outermost concentric circle, the extracted feature values expressed by a two-dimensional histogram for the distance and angle from the center of the concentric circles.

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

identification means for identifying the subject of the input image with the subject of the model image when the number of matching pairs, which is the result of the operation for determining whether the input image feature value matches any of the plurality of model image feature values, is greater than a predetermined value.

8. An information processing method used in an information processing apparatus that compares an input image with a model image to identify the subject of the input image with the subject of the model image, the method comprising the steps of:

setting feature points, each of which is on an edge of the model image and provided to extract a model image feature value, which is the feature value of the model image, and extracting the model image feature value from each of a plurality of feature value extraction areas in the neighborhood of each of the feature points; and

determining whether an input image feature value, which is the feature value of the input image, at the point that is on an edge of the input image and corresponds to a feature point matches any of the plurality of model image feature values at the feature point,

wherein the step of setting feature points selects a plurality of edge points as feature points based on an entropy value indicating how the plurality of edge points are scattered.

9. A non-tranitory computer readable medium storing a program that causes a computer to carry out a process for comparing an input image with a model image to identify the subject of the input image with the subject of the model image, the process comprising the steps of:

setting feature points, each of which is on an edge of the model image and provided to extract a model image feature value, which is the feature value of the model image, and extracting the model image feature value from each of a plurality of feature value extraction areas in the neighborhood of each of the feature points; and

determining whether an input image feature value, which is the feature value of the input image, at the point that is on an edge of the input image and corresponds to a feature point matches any of the plurality of model image feature values at the feature point,

wherein the step of setting feature points selects a plurality of edge points as feature points based on an entropy value indicating how the plurality of edge points are scattered.

10. An information processing apparatus that compares an input image with a model image to identify the subject of the input image with the subject of the model image, the apparatus comprising:

a feature value extractor configured to set feature points, each of which is on an edge of the model image and provided to extract a model image feature value, which is the feature value of the model image, and extracting the model image feature value from each of a plurality of feature value extraction areas in the neighborhood of each of the feature points; and

a matching section configured to determine whether an input image feature value, which is the feature value of the input image, at the point that is on an edge of the input image and corresponds to feature point matches any of the plurality of model image feature values at the feature point,

wherein the feature value extractor selects a plurality of edge points as feature points based on an entropy value indicating how the plurality of edge points are scattered.