Fast matching systems and methods for personal identification

• US 6,944,318 B1
• Filed: 01/14/2000
• Issued: 09/13/2005
• Est. Priority Date: 01/15/1999
• Status: Expired due to Term

• Current Assignee: Citicorp Credit Services Incorporated (Citigroup Incorporated)
• Original Assignee: Citicorp Development Center Incorporated (Citigroup Incorporated)
• Inventors: Khan, Mohammed A., Takata, Melvin M., Merman, Michael
• Primary Examiner(s): AHMED, SAMIR ANWAR
• Application Number:
  US09/483,561
• Time in Patent Office: 2,069 Days
• Field of Search: 382/115, 382/116, 382/117, 382/118, 235/380, 340/5.82, 340/5.83, 704/246, 704/273, 902/4, 902/25, 902/3, 351/200, 351/204, 359/2
• US Class Current: 382/115
• CPC Class Codes: G06V 40/197 Matching; Classification

Abstract

Systems and methods for personal identification, preferably based on iris identification, include a camera for acquiring a presented image of a presented individual and a computer device for converting the presented image into a presented image metric. The computer device also compares the presented image metric to a predetermined arrangement of a plurality of reference image metrics of known individuals to identify the presented individual as one of the known individuals. The presented image metric and the plurality of reference metrics are both formatted to represent detailed physical characteristics of the individuals, including iris characteristics and other physical characteristics. Further, the plurality of reference image metrics may also be formatted to represent user-defined characteristics, which include non-image-related characteristics. The predetermined arrangement of the plurality of reference image metrics are ordered to provide a fast match with the presented image metric. As such, the predetermined arrangement may be based on iris characteristics, other physical characteristics and user-defined characteristics. Alternately, the system of the present invention may include a carousel system that presents a circulating, predetermined arrangement of a series of reference image metrics that are compared with a plurality of presented image metrics in parallel. Thus, the systems and methods provide a fast match determination between the presented image metric and the plurality of reference image metrics in order to identify the presented individual.

Claims

1. A computer system for identifying a presented individual represented at least in part by presented image data, comprising:

a memory having a plurality of reference metrics, wherein each of the plurality of reference metrics represents a reference image of at least a portion of one of a plurality of known individuals, and wherein each of the plurality of reference metrics includes at least one classifiable reference characteristic of the reference image;

a processor in communication with the memory and operable to receive the presented image data, the processor operable to run a program to convert the presented image data to a presented image metric having at least one classifiable characteristic of the presented individual, the program further operable to retrieve at least a predetermined arrangement of the plurality of reference metrics, wherein the program generates an identification signal to identify the presented individual as one of the plurality of known individuals if a match is found between the presented image metric and one of the predetermined arrangement of reference metrics.

2. The system of claim 1, wherein the predetermined arrangement is based on the at least one classifiable characteristic of the presented image metric.

3. The system of claim 2, wherein the at least one classifiable characteristic of the presented image metric comprises a physical characteristic of the presented individual.

4. The system of claim 3, wherein the physical characteristic is selected from a group consisting of a facial characteristic, hair color, skin tone and an iris characteristic.

5. The system of claim 2, wherein the at least one classifiable characteristic of the presented image metric comprises a characteristic of a presented iris of the presented individual.

6. The system of claim 1, wherein the predetermined arrangement is based on a user-defined characteristic of each of the plurality of reference metrics.

7. The system of claim 6 wherein the user-defined characteristic comprises a non-image-related characteristic of each of the plurality of known individuals.

8. The system of claim 7, wherein the non-image-related characteristic is selected from the group consisting of a geographical characteristic, an affiliation characteristic and an individual habit characteristic.

9. The system of claim 1, wherein the predetermined arrangement is a selected subset of the plurality of reference metrics based on an ordered search of the plurality of reference metrics.

10. The system of claim 1, wherein the at least one classifiable characteristic of the reference image includes a non-image-related characteristic of each of the plurality of known individuals, and wherein the predetermined arrangement is based on a combination of the at least one classifiable characteristic of the presented image metric and the non-image-related characteristic of each of the plurality of known individuals.

11. The system of claim 1, wherein the predetermined arrangement is based on a combination of a characteristic of a presented iris of the presented individual, another one of a physical characteristic of the presented individual and a non-image-related characteristic of each of the plurality of known individuals.

12. The system of claim 1, wherein the predetermined arrangement comprises binning the plurality of reference metrics based on the at least one classifiable reference characteristic of the plurality of known individuals such that reference metrics having similar reference characteristics are arranged in the same bin.

13. The system of claim 1, further comprising a carousel program within the processor, wherein the carousel program is operable for concurrently determining a match between a plurality of received presented image metrics and the plurality of reference metrics, wherein each of the plurality of presented image metrics represents at least one characteristic of a presented image of one of a plurality of presented individuals, and wherein the predetermined arrangement comprises a circular presentation of the plurality of reference metrics, and wherein the carousel program generate an identification signal to identify each of the plurality of presented individuals as one of the plurality of known individuals if a match is found between one of the plurality of presented image metrics and one of the plurality of reference metrics.

14. The system of claim 1, further comprising a camera operable for capturing the presented image of the presented individual and outputting the presented image data representing the presented image.

15. The system of claim 1, further comprising a secure system having restricted access, wherein the identification signal is receivable by the secure system for determining access.