Face recognition with combined PCA-based datasets

• US 8,897,503 B2
• Filed: 09/12/2011
• Issued: 11/25/2014
• Est. Priority Date: 08/02/2006
• Status: Active Grant

• Current Assignee: FotoNation Limited (Adeia, Inc.), National University of Ireland Galway
• Original Assignee: DigitalOptics Corporation Europe Limited (Adeia, Inc.)
• Inventors: Corcoran, Peter, Costache, Gabriel N.
• Primary Examiner(s): STREGE, JOHN B
• Application Number:
  US13/230,768
• Publication Number:
  US 20120230554A1
• Time in Patent Office: 1,170 Days
• Field of Search: 382/118, 382/224
• US Class Current: 382/118
• CPC Class Codes: G06F 18/2135 based on approximation criteria, e.g. principal component analysis
• CPC Class Codes: G06F 18/253 of extracted features
• CPC Class Codes: G06V 10/7715 Feature extraction, e.g. by transforming the feature space, e.g. multi-dimensional scaling [MDS]; Mappings, e.g. subspace methods
• CPC Class Codes: G06V 10/806 of extracted features
• CPC Class Codes: G06V 40/169 Holistic features and representations, i.e. based on the facial image taken as a whole

Abstract

A representation framework is determined in a face recognition method for a first collection of facial images including at least principle component analysis (PCA) features. A representation of said first collection is stored using the representation framework. A modified representation framework is determined based on statistical properties of original facial image samples of a second collection of facial images and the stored representation of the first collection. The first and second collections are combined without using original facial image samples. A representation of the combined image collection (super-collection) is stored using the modified representation framework.

Claims

1. A consumer digital camera including a processor programmed by digital code to perform a face recognition method for working with two or more collections of facial images, wherein the method comprises:

(a) determining distinct representation frameworks for first and second sets of facial regions extracted from corresponding image collections each framework including at least principle component analysis (PCA) features;

(b) storing representations of said first and second sets of facial regions using their respective representation frameworks;

(c) determining a third, distinct representation framework based on the representations of the first and second sets of facial regions;

(d) combining the stored representation of the first and the second sets of facial regions without using original facial image samples;

including back-projecting the sets of facial regions into their respective representation frameworks, combining the two back-projected representations of these sets of facial regions into a combined dataset, and forward projecting the combined dataset into the third representation framework;

(e) storing a representation of the combined dataset using said third representation framework;

(f) comparing a representation of a current facial region, determined in terms of said third representation framework, with one or more representations of facial images of the combined dataset; and

(g) based on the comparing, determining whether one or more of the facial images within the combined dataset matches the current facial image.

2. The consumer digital camera of claim 1, wherein the method further comprises updating the PCA features based on first eigenvectors and first sets of eigenvalues for each of the first and second sets of facial regions.

3. The consumer digital camera of claim 2, wherein the updating comprises calculating a new set of eigenvectors from the previously calculated first eigenvectors of the first and second sets of facial regions.

4. The consumer digital camera of claim 1, wherein the method does not use original data samples of the first and second sets of facial regions.

5. The consumer digital camera of claim 1, wherein the first and second sets of facial regions contain samples of different dimension, and the method further comprises:

(i) choosing a standard size of face region, and

(ii) re-sizing eigenvectors using interpolation to the standard size.

6. The consumer digital camera of claim 5, wherein said samples of different dimension were analyzed using different standard sizes of face region.

7. The consumer digital camera of claim 1, wherein the method further comprises generating the third representation framework according to the following: