Method and apparatus of recognizing face using component-based 2nd-order principal component analysis (PCA)/independent component analysis (ICA)

• US 7,254,257 B2
• Filed: 03/04/2003
• Issued: 08/07/2007
• Est. Priority Date: 03/04/2002
• Status: Active Grant

• Current Assignee: Samsung Electronics Co. Ltd.
• Original Assignee: Samsung Electronics Co. Ltd.
• Inventors: Lee, Jong-ha, Kim, Tae-kyun, Kim, Hyun-woo, Kee, Seok-cheol, Hwang, Won-jun
• Primary Examiner(s): MARIAM, DANIEL G
• Application Number:
  US10/377,748
• Publication Number:
  US 20030169908A1
• Time in Patent Office: 1,617 Days
• Field of Search: 382/118, 382/160, 382/181, 382/190, 382/209, 382/218, 382/228, 382/305, 382/173, 707/3, 707/6
• US Class Current: 382/118
• CPC Class Codes: G06F 18/21342 using statistical independence, i.e. minimising mutual information or maximising non-gaussianity
• CPC Class Codes: G06F 18/2135 based on approximation criteria, e.g. principal component analysis
• 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 40/171 Local features and components; Facial parts eye characteristics G06V40/18; Occluding parts, e.g. glasses; Geometrical relationships

Abstract

A method and apparatus for recognizing and searching for a face using 2nd-order independent component analysis (ICA) are provided. The method for describing feature points uses 2nd-order ICA d to describe a facial image space and improve recognition performance in various illumination conditions. According to the method and apparatus, use of pose or illumination invariant face descriptor enables retrieval of human faces and authentication of a specific individual.

Claims

1. A method of recognizing a face using component-based independent component analysis (ICA)/principal component analysis (PCA), the method comprising the steps of:

(a) segmenting a predetermined input facial image by facial components and generating segmented images;

(b) performing PCA on the segmented images to generate low-pass filtered images;

(c) calculating a difference between the low-pass filtered images and the segmented images to obtain residual images; and

(d) performing ICA on the residual images and generating first feature information corresponding to the segmented images, and generating pieces of second feature information as a set of the pieces of the first feature information.

2. A computer-readable recording medium in which a program for executing the method claimed in claim 1 in a computer, is recorded.

3. A method of recognizing a face using component-based principal component analysis (PCA)/independent component analysis (ICA), the method comprising the steps of:

(a) receiving a first facial image to be recognized and a second facial image to be used as a basis for comparison with the first facial image;

(b) performing component-based PCA/ICA on the first and second facial images, respectively, and extracting pieces of second feature information corresponding to the first and second facial images, wherein the performing component-based PCA/ICA includes performing PCA on the first and second facial images, obtaining residual images by calculating a difference between images generated by the PCA and the first and second facial images, and performing ICA on the residual images; and

(c) determining whether there is a similarity between the pieces of the second feature information corresponding to the first and second facial images, and outputting the determination result.

4. The method of claim 3, wherein the step (b) comprises:

(b1) segmenting the first and second facial images by facial components and generating first and second segmented images;

(b2) performing PCA on the first and second segmented images to generate first and second low-pass filtered images corresponding thereto by removing high-frequency components from the first and second segmented images;

(b3) calculating a difference between the low-pass filtered images and the first and second segmented images to obtain first and second residual images having high-frequency components; and

(b4) performing ICA on the first and second residual images to generate pieces of first feature information corresponding to the first and second segmented images, and generating pieces of second feature information as sets of pieces of the corresponding first feature information.

5. The method of claim 3, wherein the step (c) comprises:

(c1) expressing the pieces of first feature information corresponding to the first and second segmented images as vectors;

(c2) obtaining a first result values by dividing an inner product of a pair of vectors corresponding to same components of the first and second facial images by norms of the vectors; and

(c3) obtaining a sum of the first result values and determining whether the first and second facial images identify a same person by comparing the sum to a predetermined reference value.

6. The method of claim 5, wherein the step (c3) comprises obtaining a sum of second result values multiplied by a predetermined weighting factor and determining whether the first and second facial images identify the same person by comparing the sum to a predetermined reference value.

7. A computer-readable recording medium in which a program for executing the method claimed in claim 3 in a computer, is recorded.

8. A method of searching for a face using component-based principal component analysis (PCA)/independent component analysis (ICA), the method comprising:

(a) constructing a database (DB) in which facial images of one or more persons and identification information of the persons are stored;

(b) receiving a facial image to be searched for from the DB;

(c) performing the component-based PCA/ICA on the facial image to be searched for and generating second feature information, wherein the performing component-based PCA/ICA includes performing PCA on the facial image to be searched for, obtaining a residual image by calculating a difference between an image generated by the PCA and the facial image to be searched for, and performing ICA on the residual image;

(d) sequentially reading facial images stored in the DB, and performing the component-based PCA/ICA on the read facial images and generating second feature information;

(e) comparing the second feature information corresponding to a facial image read from the DB with the second feature information corresponding to the facial image to be searched for, and determining whether the two facial images identify a same person; and

(f) reading identification information corresponding to the facial image from the DB and outputting the identification information if it is determined that the facial images identify the same person.

9. The method of claim 8, wherein the generating the second feature information of step (c) comprises:

segmenting the facial images by facial components and generating segmented facial images;

performing PCA on the segmented facial images to generate low-pass filtered images by removing high-frequency components from the segmented facial images;

calculating a difference between the low-pass filtered images and the segmented images to generate residual images corresponding to the segmented facial images; and

performing ICA on the residual images which correspond to the segmented facial images and generating pieces of first feature information as ICA coefficient matrices, and generating pieces of second feature information as sets of pieces of the corresponding first feature information.

10. The method of claim 8, wherein the step (e) comprises:

(e1) expressing the pieces of first feature information corresponding to the facial image to be searched for and the facial image stored in the DB as vectors;

(e2) obtaining a first result values by dividing an inner product of a pair of vectors corresponding to same components of the facial images by norms of the vectors; and

(e3) obtaining a sum of the first result values and determining whether the facial images identify the same person by comparing the sum to a predetermined reference value.

11. The method of claim 10, wherein the step (e3) comprises obtaining second result values by multiplying respective first result values by a predetermined weighting factor, and determining whether the facial images identify the same person by comparing the sum to a predetermined reference value.

12. A computer-readable recording medium in which a program for executing the method claimed in claim 8 in a computer, is recorded.

13. A method of searching for a face using component-based principal component analysis (PCA)/independent component analysis (ICA), the method comprising:

(a) constructing a database (DB) in which facial images of one or more persons and identification information of the persons are stored;

(b) receiving a facial image to be searched for from the DB;

(c) performing the component-based PCA/ICA on the facial image to be searched for and generating second feature information, wherein the performing component-based PCA/ICA includes performing PCA on the facial image to be searched for, obtaining a residual image by calculating a difference between an image generated by the PCA and the facial image to be searched for, and performing ICA on the residual image; and

(d) detecting second feature information similar to the second feature information corresponding to the facial image to be searched for from the DB and outputting the detected second feature information.

14. A computer-readable recording medium in which a program for executing the method claimed in claim 13 in a computer, is recorded.

15. An apparatus for generating feature information using component-based principal component analysis (PCA)/independent component analysis (ICA), the apparatus comprising:

an image segmenting portion which segments a predetermined input original facial image by facial components included in the input original facial image and generates segmented facial images;

a low-pass filtered image generator which performs PCA on the segmented facial images to generate low-pass filtered images by removing high-frequency components from the input image;

a residual image generator which calculates a difference between the low-pass filtered images and the segmented images to generate residual images; and

an ICA coefficient matrix calculator which performs ICA on the residual images and generates pieces of first feature information corresponding to the segmented images, and generates pieces of second feature information as a set of the pieces of first feature information.

16. An apparatus for extracting a face using component-based principal component analysis (PCA)/independent component analysis (ICA), the apparatus comprising:

an input portion which receives a first facial image to be recognized and a second facial image to be used as a basis for comparison with the first facial image;

a feature information generator which performs component-based PCA/ICA on the first and second facial images and generates pieces of second feature information as a set of ICA coefficient matrices corresponding to facial components, wherein the feature information generator performs PCA on the first and second facial images, obtains residual images by calculating a difference between images generated by the PCA and the first and second facial images, and performs ICA on the residual images; and

a comparator which compares a piece of second feature information corresponding to the first facial image to a piece of second feature information corresponding to the second facial image to determine whether there is a similarity therebetween, and outputs the determination result.

17. The apparatus of claim 16, wherein the feature information generator comprises:

an image segmenting portion which segments the input original facial image by facial components and generates segmented facial images;

a low-pass filtering portion which performs PCA on the segmented images to generate low-pass filtered images by removing high-frequency components from the segmented images;

a residual image generator which calculates a difference between the low-pass filtered images and the segmented images to generate residual images having the high-frequency components; and

an ICA coefficient matrix calculator which performs ICA on the residual images and generates pieces of first feature information corresponding to the segmented images, and generates pieces of second feature information as a set of the pieces of first feature information.

18. An apparatus for searching for a face using component-based principal component analysis (PCA)/independent component analysis (ICA), the apparatus comprising:

an input portion which receives facial images used in constructing a facial image information database (DB) and a facial image to be searched for that is included in the DB;

a feature information generator which receives facial images from the input portion, performs the component-based PCA/ICA on the facial images and generates pieces of second feature information corresponding to the facial images, wherein the feature information generator performs PCA on the facial images, obtains residual images by calculating a difference between images generated by the PCA and the facial images, and performs ICA on the residual images;

a DB managing portion which stores the second feature information corresponding to the facial images used in constructing the DB and supplied from the feature information generator, and manages the information to be stored in the DB; and

a comparator which receives the second feature information corresponding to the facial image to be searched for from the feature information generator, searches for second feature information similar to the received second feature information, and determines whether the facial images identify a same person by comparing a degree of similarity to a predetermined reference value.

19. The apparatus of claim 18, wherein the DB managing portion receives and stores the facial images used in constructing the DB from the input portion,

wherein the comparator sequentially reads facial images stored in the DB and supplies the read facial images to the feature information generator, receives pieces of second feature information corresponding to a facial image in the DB, compares the second feature information corresponding to the facial image stored in the DB with the second feature information corresponding to the facial image to be searched for, and determines whether the facial images identify the same person, and

wherein the feature information generator receives the facial image to be searched for from the input portion, receives the facial image in the DB from the comparator, performs the component-based PCA/ICA on the facial images and generates pieces of second feature information corresponding to the facial images.

20. The apparatus of claim 18, wherein the feature information generator comprises:

an image segmenting portion which segments the facial images used in constructing the DB and the facial image to be searched for by facial components and generates segmented facial images;

a low-pass filtering portion which performs PCA on the segmented facial images to generate low-pass filtered images by removing high-frequency components from the segmented facial images;

a residual image generator which calculates a difference between the low-pass filtered images and the segmented images to generate residual images having the high-frequency components; and

an ICA coefficient matrix calculator which performs ICA on the residual images and generates pieces of first feature information corresponding to the segmented images, and generates pieces of second feature information as a set of the pieces of first feature information.