Face recognition method
First Claim
1. A face recognition method comprising:
- storing a plurality of aggregations of sample feature data matched with corresponding personnel data, each of the plurality of aggregations of sample feature data including a ratio of a plurality of widths of feature areas of a sample face;
receiving, from an imager, an input image of a person;
extracting, from the input image, input feature data that are each equal to or more than a critical value;
comparing an aggregation of the input feature data with each of the plurality of stored aggregations;
selecting an aggregation among the plurality of stored aggregations having the greatest similarity with the aggregation of the input feature data; and
identifying the person based on the personnel data matched with the selected aggregation.
1 Assignment
0 Petitions
Accused Products
Abstract
A face recognition method of the present disclosure includes configuring aggregations of feature data that include a plurality of feature data of faces and match to a plurality of personnel data; extracting from an input image a plurality of input feature data that correspond to the feature data and that is equal to or more than a critical value; comparing an aggregation of input feature data that includes the input feature data with each of the pre-stored aggregations of feature data, and selecting from the aggregations of feature data an aggregation of feature data having the greatest similarity with the aggregation of the input feature data; and identifying a person on the image based on personnel data that matches the aggregation of feature data having the greatest similarity.
14 Citations
20 Claims
-
1. A face recognition method comprising:
-
storing a plurality of aggregations of sample feature data matched with corresponding personnel data, each of the plurality of aggregations of sample feature data including a ratio of a plurality of widths of feature areas of a sample face; receiving, from an imager, an input image of a person; extracting, from the input image, input feature data that are each equal to or more than a critical value; comparing an aggregation of the input feature data with each of the plurality of stored aggregations; selecting an aggregation among the plurality of stored aggregations having the greatest similarity with the aggregation of the input feature data; and identifying the person based on the personnel data matched with the selected aggregation. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18)
-
-
19. A face recognition method comprising:
-
pre-storing aggregations of feature data that include a plurality of feature data of faces and match to a plurality of personnel data; receiving, from an imager, an input image of a person; extracting, from the input image, a plurality of input feature data that correspond to the feature data and that is equal to or more than a critical value; comparing an aggregation of input feature data that includes the input feature data with each of the pre-stored aggregations of feature data, and selecting from the aggregations of feature data an aggregation of feature data having the greatest similarity with the aggregation of the input feature data; and identifying the person based on personnel data that matches the aggregation of feature data having the greatest similarity, wherein the feature data comprises; first feature data that is a ratio of a width of a right eye and a first face width; second feature data that is a ratio of a width of a nose and a second face width; third feature data that is a ratio of a width of a left eye and the width of the nose; fourth feature that is a ratio of a width of a mouth and a third face width; fifth feature that is a ratio of the width of the right eye and the width of the mouth; and sixth feature data that is a ratio of the width of the nose and the width of the mouth. - View Dependent Claims (20)
-
Specification