PATTERN RECOGNITION METHOD AND PATTERN RECOGNITION APPARATUS

• US 20110158540A1
• Filed: 12/16/2010
• Published: 06/30/2011
• Est. Priority Date: 12/24/2009
• Status: Active Grant

• Current Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Original Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Inventors: Suzuki, Takashi, Sato, Hiroshi, Mori, Katsuhiko
• Granted Patent: US 9,092,662 B2
• Time in Patent Office: Days
• US Class Current: 382/195
• CPC Class Codes: G06V 40/171 Local features and components; Facial parts eye characteristics G06V40/18; Occluding parts, e.g. glasses; Geometrical relationships
• CPC Class Codes: G06V 40/172 Classification, e.g. identification

Abstract

A pattern recognition apparatus that recognizes a data attribute of input data calculates correlation values of feature quantities of corresponding local patterns between the input data and dictionary data for each of a plurality of dictionary data prepared for each data attribute, combines, for each data attribute, the calculated correlation values of local patterns of each dictionary datum to acquire a set of correlation values of each data attribute, integrates correlation values included in each set of correlation values of each data attribute to calculate a similarity of the input data for each data attribute, and identifies the data attribute of the input data based on the calculated similarity.

Claims

1. A pattern recognition apparatus that recognizes a data attribute of input data, the apparatus comprising:

a correlation value calculation unit configured to calculate correlation values of feature quantities of corresponding local patterns between the input data and dictionary data for each of a plurality of dictionary data prepared for each data attribute and configured to combine, for each data attribute, the calculated correlation values of local patterns of each dictionary datum to acquire a set of correlation values of each data attribute;

a similarity calculation unit configured to integrate correlation values included in each set of correlation values of each data attribute to calculate a similarity of the input data for each data attribute; and

an identification unit configured to identify the data attribute of the input data based on the similarity calculated by said similarity calculation unit.

2. The apparatus according to claim 1, wherein

said similarity calculation unit calculates an average value of only the correlation values greater than a threshold, a weighted average value of the correlation values, or a maximum value of the correlation values for each set of correlation values of each data attribute to set the similarity of the data attribute.

3. The apparatus according to claim 1, wherein

said similarity calculation unit integrates the correlation values of corresponding local patterns for each set of correlation values of each data attribute to acquire a set of combined correlation values and integrates the acquired combined correlation values to calculate a similarity of the input data for each data attribute.

4. The apparatus according to claim 1, further comprising:

an estimation unit configured to estimate a variation pattern in the input data; and

a determination unit configured to determine the local pattern used for each dictionary datum based on the variation pattern of the input data and a variation pattern of each dictionary datum, wherein

said similarity calculation unit calculates the similarity of the input data for each data attribute by integrating the set of correlation values of each dictionary datum corresponding to the local pattern to be used, the set included in each set of correlation values of each data attribute, the set determined by said determination unit.

5. The apparatus according to claim 4, wherein

said similarity calculation unit integrates the correlation values for each local pattern to be used in each set of correlation values for each data attribute to acquire a set of combined correlation values and integrates the acquired combined correlation values to calculate the similarity of the input data for each data attribute.

6. The apparatus according to claim 3, wherein

the combined correlation value is an average value of only the correlation values over the threshold, a weighted average value of the correlation values, or a maximum value of the correlation values.

7. The apparatus according to claim 3, wherein

said similarity calculation unit calculates an average value of only the combined correlation values over the threshold, a weighted average value of the combined correlation values, or a maximum value of the combined correlation values for each set of combined correlation values of each data attribute to set the similarity of the data attribute.

8. The apparatus according to claim 4, wherein

the input data is image data, and

the variation pattern is at least one of variations in face orientation of a subject included in the image, variations in illumination, variations in expression, and occlusions of the face.

9. The apparatus according to claim 1, wherein

the data attribute is individual identification information.

10. A pattern recognition method by a pattern recognition apparatus that recognizes a data attribute of input data, the method comprising:

a correlation value calculation step of calculating correlation values of feature quantities of corresponding local patterns between the input data and dictionary data for each of a plurality of dictionary data prepared for each data attribute and combining, for each data attribute, the calculated correlation values of local patterns of each dictionary datum to acquire a set of correlation values of each data attribute;

a similarity calculation step of integrating the set of correlation values included in each set of correlation values of each data attribute to calculate a similarity of the input data for each data attribute; and

an identification step of identifying the data attribute of the input data based on the similarity calculated in said similarity calculation step.

11. A computer-readable storage medium storing a computer program causing a computer to execute the steps of the pattern recognition method according to claim 10.