Feature extraction system and face image recognition system

• US 6,430,307 B1
• Filed: 04/13/1999
• Issued: 08/06/2002
• Est. Priority Date: 06/18/1996
• Status: Expired due to Fees

• Current Assignee: Panasonic Intellectual Property Corporation of America (Panasonic Holdings Corporation)
• Original Assignee: Matsushita Electric Industrial Company Limited (Panasonic Holdings Corporation)
• Inventors: Souma, Masaki, Nagao, Kenji
• Primary Examiner(s): Tran, Phuoc
• Application Number:
  US09/290,256
• Time in Patent Office: 1,211 Days
• Field of Search: 382/118, 382/160, 382/190, 382/219, 382/220, 382/224, 382/225, 382/228
• US Class Current: 382/118
• CPC Class Codes: G06F 18/213 Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
• CPC Class Codes: G06F 40/16 Automatic learning of transformation rules, e.g. from examples
• CPC Class Codes: G06V 40/16 Human faces, e.g. facial parts, sketches or expressions

Abstract

A feature extraction and pattern collation system for statistically analyzing a set of samples of feature vectors to calculate a feature being an index for a pattern identification, which is capable of identifying confusing data with a high robustness. In this system, a storage section stores a feature vector inputted through an input section and a neighborhood vector selection section selects a specific feature vector from the feature vectors existing in the storage section. The specific feature is a neighborhood vector close in distance to the feature vector stored in the storage section. Further, the system is equipped with a feature vector space production section for outputting a partial vector space. The partial vector space is made to maximize the local scattering property of the feature vector when the feature vector is orthogonally projected to that space. The pattern collation system is thus capable of obtaining learning data from a plurlaity of different persons.

Claims

1. A pattern collation system comprising:

a first sample pattern input means for receiving a set of samples of face patterns taken under a first photographing condition;

a second sample pattern input means for receiving a set of samples of face patterns of a person identical to a person of said face patterns inputted through said first sample pattern input means, with said face patterns being taken under a second photographing condition;

first storage means for storing said pattern sample set inputted through said first sample pattern input means, second storage means for storing said pattern sample set inputted through said second sample pattern input means;

a difference pattern producing means for repeatedly deriving each of patterns corresponding to the identical person from each of said pattern sample sets stored in said first and second storage means to calculate a pattern corresponding to a difference therebetween;

a principal-component analyzing means for making a principal-component analysis of a set of patterns obtained by addition of said pattern sample set stored in said first storage means and said difference pattern produced by said difference pattern producing means to output a principal component and a value expressing a variance in a principal-component direction;

an analysis result storage means for storing data outputted from said principal-component analyzing means;

a collation pattern input means for receiving two patterns to be collated; and

a similarity calculating means for referring to said data stored in said analysis result storage means to calculate, on the basis of that result, a degree of similarity between said patterns read out from said collation pattern input means by weighting a component of said patterns in said principal-component direction.