Method and system for indexing and searching an iris image database

• US 7,761,453 B2
• Filed: 03/02/2007
• Issued: 07/20/2010
• Est. Priority Date: 01/26/2005
• Status: Active Grant

• Current Assignee: Gentex Corporation
• Original Assignee: Honeywell International Inc.
• Inventors: Hamza, Rida M.
• Primary Examiner(s): Vo, Tim T.
• Assistant Examiner(s): Fan, Shiow-Jy
• Application Number:
  US11/681,751
• Publication Number:
  US 20070276853A1
• Time in Patent Office: 1,236 Days
• Field of Search: 707/2, 707100-101, 707/736, 707/790, 707/791, 707/802, 707/E17.039
• US Class Current: 707/737
• CPC Class Codes: G06F 16/22 Indexing; Data structures therefor; Storage structures
• CPC Class Codes: G06F 16/90344 by using string matching techniques
• CPC Class Codes: G06F 18/24323 Tree-organised classifiers
• CPC Class Codes: G06V 10/764 using classification, e.g. of video objects
• CPC Class Codes: G06V 40/197 Matching; Classification

Abstract

A method and system for indexing and searching a database of iris images having a system to expedite a process of matching a subject to millions (more or less) of templates within a database is disclosed. The system may progressively match an iris image template to an iris template in a database by progressing from a top layer of the database to a lower layer of the database. Such matching or retrieval may use a subject code as a query or probe and then find a similarity measure for the features of codes or templates in the database. A multi-stage hierarchal clustering process may be used to compress codes and/or templates.

Claims

1. A method of indexing and searching a database comprising:

providing a storage module;

providing a comparator module communicatively coupled to the storage module;

providing a database of iris templates within the storage module, wherein each iris template in the database comprises a barcode configured to have a certain number of bits;

hierarchically partitioning the database into a plurality of layers of sub-databases of the iris templates based on phase information, radial information, and/or angular information of an iris image, wherein each iris template in each of the layers of the sub-databases comprises a segment of the iris template barcode and the sum of the iris templates of each layer of sub-databases represents all the iris templates in the database;

providing a barcode of an iris image for matching to an iris template in the database, wherein the barcode of the iris image has the same configuration as the configuration of each iris template in the database;

selecting a segment of the barcode of the iris image that corresponds to a segment of an iris template in a layer of the sub-databases; and

progressively matching the selected segment of the barcode of the iris image to an iris template in a layer of the sub-databases by progressing from a top layer of the sub-databases to a lower layer of the sub-databases; and

wherein the layers of the sub-databases comprise;

phase clustering;

radial clustering;

angular clustering; and

bit clustering or binning.

2. The method of claim 1, wherein the segment of the iris template barcode includes only a combination or transformation of a portion of the bits in a barcode.

3. The method of claim 1, wherein hierarchically partitioning the database into a plurality of sub-databases of templates based on phase information, radial information, and/or angular information includes two or more types of information.

4. The method of claim 3, wherein the two or more types of information is progressive relative to the number of layers of the sub-databases in the plurality of sub-databases and a segment of each iris template in the sub-database and the segment is a number of bits representing an iris template of the sub-database.

5. The method of claim 4, wherein a first layer of the sub-databases of the database has iris templates represented by an iris template including a larger number of bits than a second layer of the sub-databases.

6. The method of claim 5, wherein a number of iris templates in the first layer of the sub-databases is the same as or larger than a number of iris templates in the second layer of the sub-databases.

7. The method of claim 1, wherein a number of iris templates used for matching is reduced when progressing from one layer of the sub-databases to another layer of the sub-databases.

8. The method of claim 6, wherein a matching of an iris image barcode is reduced to a least number of iris templates.

9. The method of claim 8, wherein bit summation or combination or transformation of bits on a barcode is used to compress the barcode.

10. The method of claim 9, wherein bit summation or transformation on a barcode bits is used to increase a speed of matching an iris image barcode to an iris template in the database.

11. The method of claim 8, wherein bit summation on a horizontal or vertical axis on the barcode is used to increase a speed of matching the barcode to an iris template in the database.

12. The method of claim 1, wherein the radial information to start is based on information close to an inner border of an iris closest to the pupil borders.

13. The method of claim 1, wherein the angular information is based on angular segments and sclera-iris angular segments away from the eyelids are considered for matching before other segments.

14. A method for matching an iris image barcode to a database comprising:

providing a barcode of an iris image;

providing a storage module;

providing a comparator module communicatively coupled to the storage module;

providing a database comprising a certain number of iris templates within the storage module, the iris templates configured to have a certain number of bits;

hierarchically partitioning the database into multiple layers of datasets based on phase information, radial information, and/or angular information, the number of iris templates in each dataset are equal to or smaller than the number of iris templates in the database and the certain number of templates in each dataset configured to have a segment of the iris template as a certain number of bits equal to or smaller than the certain number of bits of the iris templates in the database;

matching a segment of the barcode of an iris image to a corresponding segment of an iris template stored in the database by progressively matching the certain number of bits of the iris image barcode to the corresponding certain number of bits of the iris templates in the dataset progressing from one layer of the dataset to another layer of the dataset.

15. The method of claim 14 wherein:

the matching of the segment of the barcode of an iris image to an iris template of the dataset results in a number of iris templates which are a possible match for the barcode of an iris image; and

the number of iris templates which are a possible match for the barcode of an iris image is less than the number of iris templates in the database.

16. A method for searching a database comprising:

providing a storage module;

providing a comparator module communicatively coupled to the storage module;

providing a database comprising iris templates within the storage module, the iris templates configured to include information as a certain number of bits;

providing an iris image barcode configured to include information as a certain number of bits corresponding to the certain number of bits of the iris templates in the database;

partitioning the database into a number of sub-databases based on phase information, the number of sub-databases comprising a number of iris templates, the number of iris templates configured to include information as a segment of the iris template having a certain number of bits;

partitioning the sub-databases into a number of datasets based on angular information, the number of datasets comprising a number of iris templates, the number of iris templates configured to include information as a segment of the iris template having a certain number of bits;

partitioning the datasets into a number of sub-datasets based on radial information, the number of sub-datasets comprising a number of iris templates, the number of iris templates configured to include information as a segment of the iris template having a certain number of bits; and

matching a segment of the iris barcode to a corresponding segment of an iris template in the database by progressively matching the certain number of bits of the iris image barcode to the corresponding certain number of bits of the iris templates progressing from the sub-databases to the datasets, and to the sub-datasets;

wherein;

the sum of the iris templates in all of the sub-databases represents all the iris templates in the database;

the sum of the iris templates in all of the datasets represents all the iris templates in the database;

the sum of the iris templates in all of the sub-datasets represents all the iris templates in the database;

the certain number of bits of the iris templates in the database is greater than the certain number of bits of the iris templates in the sub-databases;

the certain number of bits of the iris templates in the sub-databases is greater than the certain number of bits of the iris templates in the datasets;

the certain number of bits of the iris templates in the datasets is greater than the certain number of bits of the iris templates in the sub-datasets;

the matching of the segment of the iris image barcode to a corresponding segment of an iris templates of the sub-dataset results in a number of iris templates which are a possible match for the iris image barcode; and

the number of iris templates which are a possible match is less than the certain number of iris templates in the sub-dataset.

17. The method of claim 16, wherein a progression of templates and datasets continues until a dataset exists for all types of available information.