Dictionary learning device, pattern matching apparatus, method for learning dictionary and storage medium

Patent metadata

Field	Value
Publication number	US-9262694-B2
Application number	US-201214364467-A
Country	US
Kind code	B2
Filing date	Dec 12, 2012
Priority date	Dec 16, 2011
Publication date	Feb 16, 2016
Grant date	Feb 16, 2016

Abstract

Provided is a technology which enables further improvement of the accuracy of the determination in the pattern matching processing. A dictionary learning device 1 includes a score calculation unit 2 and a learning unit 3 . The score calculation unit 2 calculates a matching score representing a similarity-degree between a sample pattern, which is a sample of a pattern which is likely to be subjected to a pattern matching processing, and a degradation pattern resulting from a degrading processing on the sample pattern. The learning unit 3 learns a quality dictionary based on the calculated matching score and the degradation pattern. The quality dictionary is a dictionary which is used in a processing to evaluate a degradation degree (quality) of a matching target pattern of being pattern of an object on which the pattern matching processing is carried out.

First claim

The invention claimed is:

1. A dictionary learning device comprising:
a score calculation unit that calculates a matching score representing a similarity-degree between a sample pattern which is a sample of a pattern which is likely to be subjected to a pattern matching processing and a degradation pattern resulting from a degrading processing on the sample pattern; and
learning unit that, based on the calculated matching score and the degradation pattern, learns a quality dictionary which is used in a processing to evaluate a degradation degree of a matching target pattern of being target to be subjected to the pattern matching processing.

2. The dictionary learning device according to claim 1 , further comprising degrading processing unit that generates the degradation pattern by performing the degrading processing on the sample pattern.

3. The dictionary learning device according to claim 1 , further comprising feature extraction unit that extracts a feature of the sample pattern and a feature of the degradation pattern,
wherein the score calculation unit that calculates the matching score based on the extracted features of the sample pattern and the degradation pattern.

4. The dictionary learning device according to claim 1 , wherein the quality dictionary is a dictionary through which a matching score representing a similarity-degree between the matching target pattern and a virtual pattern, which relates to the matching target pattern and is in a non-degradation degree, is obtained.

5. The dictionary learning device according to claim 1 , wherein the sample pattern and the degradation pattern are each an image pattern or a speech signal pattern.

6. A pattern matching apparatus comprising:
a quality dictionary that is learned by a dictionary learning device according to claim 1 ;
quality evaluation unit that, based on the quality dictionary, evaluates a degradation degree of a matching target pattern which is target to be subjected to a pattern matching processing;
score calculation unit that calculates a matching score representing a similarity-degree among the plurality of matching target patterns by a processing corresponding to a processing through which the dictionary learning device calculates the matching score;
score compensation unit that compensates the matching score by using an evaluation result obtained by the quality evaluation unit; and
determination unit that determines a result of matching among the plurality of matching target patterns based on the compensated matching score.

7. The pattern matching apparatus according to claim 6 , further comprising feature extraction unit that extracts a feature of the matching target pattern, wherein the score calculation unit calculates the matching score based on each extracted feature of the matching target patterns.

8. A method for learning dictionary comprising:
calculating, by a computer, a matching score representing a similarity-degree between a sample pattern which is a sample of a pattern which is likely to be subjected to a pattern matching processing and a degradation pattern resulting from a degrading processing on the sample pattern; and
learning, by the computer, based on the calculated matching score and the degradation pattern, a quality dictionary which is used in a processing to evaluate a degradation degree of a matching target pattern of being target to be subjected to the pattern matching processing.

9. A non-transitory computer-readable storage medium that stores therein a control procedure which causes a computer to execute a processing, comprising the processes of:
calculating a matching score representing a similarity-degree between a sample pattern which is a sample of a pattern which is likely to be subjected to a pattern matching processing and a degradation pattern resulting from a degrading processing on the sample pattern; and
learning, based on the calculated matching score and the degradation pattern, a quality dictionary which is used in a processing to evaluate a degradation degree of a matching target pattern of being target to be subjected to the pattern matching processing.

10. A dictionary learning device comprising:
score calculation means for calculating a matching score representing a similarity-degree between a sample pattern which is a sample of a pattern which is likely to be subjected to a pattern matching processing and a degradation pattern resulting from a degrading processing on the sample pattern; and
learning means for, based on the calculated matching score and the degradation pattern, learning a quality dictionary which is used in a processing to evaluate a degradation degree of a matching target pattern of being target to be subjected to the pattern matching processing.

Assignees

• Ishii Masato
• Nec Corp

Inventors

• Ishii Masato

Classifications

• G06V10/772

  Determining representative reference patterns, e.g. averaging or distorting patterns; Generating dictionaries · CPC title

• G06V10/761

  Proximity, similarity or dissimilarity measures · CPC title

• G06F18/22

  Matching criteria, e.g. proximity measures · CPC title

• G06V10/98 Primary

  Detection or correction of errors, e.g. by rescanning the pattern or by human intervention; Evaluation of the quality of the acquired patterns · CPC title

• G06F18/28

  Determining representative reference patterns, e.g. by averaging or distorting; Generating dictionaries · CPC title