Person identification apparatus and method using concentric templates and feature point candidates
First Claim
1. A person identification apparatus, comprising:
- image input means for inputting an image of a person to be identified;
face area extraction means for extracting a face area from the inputted image of the person;
feature point extraction means for extracting a plurality of candidate of feature points from the extracted face area by superimposing a separation filter on the face area, the separation filter being a mask of concentric shapes including an outer ring and an inner ring to calculate a separability of feature value between the outer ring and the inner ring;
feature point set selection means for selecting n-sets of candidate feature points from all sets of candidate feature points according to face structure information;
feature point decision means for extracting a neighbor pattern including a position of each candidate feature point from the face area, a size of the neighbor pattern being based on a size of the separation filter, for calculating a similarity value between the neighbor pattern for each candidate feature point and a previously-registered template pattern, the template pattern being a standard image of the feature point, and for selecting one set of candidate feature points whose sum of the similarity values is highest among n-sets of the candidate feature points;
normalization means for generating a normalized face image of the face area according to the one set of candidate feature points; and
recognition means for comparing the normalized face image with previously-registered image to identify the person.
1 Assignment
0 Petitions
Accused Products
Abstract
In a person identification apparatus, a normalized face image is used to identify a person. When the face image of the person is input, a face area extraction section extracts the face area from the face image. A feature point extraction section extracts a plurality of candidates of feature points from the face area using a separation filter. A feature point set selection section selects sets of the candidates from the plurality of candidates of feature points according to face structure information. A feature point decision section extracts neighboring patterns of each candidate from the face area for the set of candidates, calculates a similarity value between the neighboring patterns of each candidate and the template pattern, and selects a subset of the candidates having the highest similarity values. A normalization section generates a normalized face image of the face area according to the selected subset of the candidates. A recognition section compares the normalized face image with the registered image to identify the person.
403 Citations
11 Claims
-
1. A person identification apparatus, comprising:
-
image input means for inputting an image of a person to be identified; face area extraction means for extracting a face area from the inputted image of the person; feature point extraction means for extracting a plurality of candidate of feature points from the extracted face area by superimposing a separation filter on the face area, the separation filter being a mask of concentric shapes including an outer ring and an inner ring to calculate a separability of feature value between the outer ring and the inner ring; feature point set selection means for selecting n-sets of candidate feature points from all sets of candidate feature points according to face structure information; feature point decision means for extracting a neighbor pattern including a position of each candidate feature point from the face area, a size of the neighbor pattern being based on a size of the separation filter, for calculating a similarity value between the neighbor pattern for each candidate feature point and a previously-registered template pattern, the template pattern being a standard image of the feature point, and for selecting one set of candidate feature points whose sum of the similarity values is highest among n-sets of the candidate feature points; normalization means for generating a normalized face image of the face area according to the one set of candidate feature points; and recognition means for comparing the normalized face image with previously-registered image to identify the person. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9)
-
-
10. A method for identifying a person corresponding to a face image, comprising the steps of:
-
inputting the face image of the person to be identified; extracting a face area from the face image of the person; extracting a plurality of candidate feature points from the face area by superimposing a separation filter on the face area, the separation filter being a mask of concentric circles including an outer ring and an inner ring to calculate a separability of feature value between the outer ring and the inner ring; selecting n-sets of candidate feature points from all sets of candidate feature points according to face structure information; extracting a neighbor pattern including a position of each candidate feature point from the face area, a size of the neighbor pattern being based on a size of the separation filter; calculating a similarity value between the neighbor pattern of each candidate feature point and a previously-registered template pattern, the template pattern being a standard image of the feature point; selecting one set of the candidate feature points whose sum of the similarity values is highest among n-sets of the candidate feature points; generating a normalized face image of the face area according to the one set of candidate feature points; and comparing the normalized face image with a previously registered image to identify the person.
-
-
11. A computer readable memory containing computer-readable instructions to identify a person, comprising:
-
instruction means for causing a computer to input a face image of the person to be identified; instruction means for causing a computer to extract a face area from the face image of the person instruction means for causing a computer to extract a plurality of candidate feature points from the face area by superimposing a separation filter on the face area, the separation filter being a mask of concentric shapes including an outer ring and an inner ring to calculate a separability of feature value between the outer ring and the inner ring; instruction means for causing a computer to select n-sets of candidate feature points from all sets of candidate feature points according to face structure information; instruction means for causing a computer to extract a neighbor pattern including a position of each candidate feature point from the face area, a size of the neighbor pattern being based on a size of the separation filter; instruction means for causing a computer to calculate a similarity value between the neighbor pattern of each candidate feature point and a previously-registered template pattern, the template pattern being a standard image of the feature point; instruction means for causing a computer to select one set of the candidate feature points whose sum of the similarity values is highest among n-sets of the candidate feature points; instruction means for causing a computer to generate a normalized face image of the face area according to the one set of candidate feature points; and instruction means for causing a computer to compare the normalized face image with a previously-registered image to identify the person.
-
Specification