Method and system for similarity-based image classification

• US 6,240,424 B1
• Filed: 04/22/1998
• Issued: 05/29/2001
• Est. Priority Date: 04/22/1998
• Status: Expired due to Term

• Current Assignee: NEC Corporation
• Original Assignee: NBC USA Incorporated
• Inventors: Hirata, Kyoji
• Primary Examiner(s): Alam, Hosain T.
• Application Number:
  US09/064,086
• Time in Patent Office: 1,133 Days
• Field of Search: 707/104, 709/217, 709/218, 709/219, 382/190, 382/192, 382/305
• US Class Current: 707/749
• CPC Class Codes: G06F 16/5846 using extracted text
• CPC Class Codes: Y10S 707/915 Image
• CPC Class Codes: Y10S 707/99931 Database or file accessing
• CPC Class Codes: Y10S 707/99945 Object-oriented database structure processing

Abstract

A method and apparatus for classifying and querying a database of images, in which the images in the database are classified using primary objects as a clustering center. A query interface is presented to the user allowing the user to use the primary objects to formulate queries to be applied to the classified database of images.

Claims

1. A method of classifying images of a database, wherein a set of primary shapes has been specified, the method comprising the steps of:

(a) calculating a similarity value between each of the images and each of the primary shapes; and

(b) classifying the images into categories based on a threshold for the similarity value, each category being represented by one of the primary shapes.

2. The method according to claim 1, wherein the primary shapes comprise base primary shapes, including a circle, a square, a triangle and a line.

3. The method according to claim 1, wherein the primary shapes comprise combinations of base primary shapes, the base primary shapes including a circle, a square, a triangle and a line.

4. The method according to claim 2, wherein the primary shapes comprise variations of the base primary shapes based upon size and aspect ratio.

5. The method according to claim 1, wherein the step of calculating the similarity value is dependent upon a location of objects within the images.

6. The method according to claim 1, wherein the step of calculating the similarity value is independent of a location of objects within the images.

7. The method according to claim 1, wherein a value of the threshold determines a number of images in the respective categories.

8. The method according to claim 1, wherein the database is a multimedia database.

9. The method according to claim 1, wherein the database is on the World Wide Web (WWW).

10. A system for classifying images of a database, wherein a set of primary shapes has been specified, the system comprising:

(a) a workstation; and

(b) an application program for calculating a similarity value between each of the images and each of the primary shapes, wherein the images are classified into categories based on a threshold for the similarity value, each category being represented by one of the primary shapes.

11. The system according to claim 10, wherein the primary shapes comprise base primary shapes, including a circle, a square, a triangle and a line.

12. The system according to claim 10, wherein the primary shapes comprise combinations of base primary shapes, the base primary shapes including a circle, a square, a triangle and a line.

13. The system according to claim 10, wherein the primary shapes comprise variations of the base primary shapes based upon size and aspect ratio.

14. The system according to claim 10, wherein, in the application program, the calculation of the similarity value is dependent upon a location of objects within the images.

15. The system according to claim 10, wherein, in the application program, the calculation of the similarity value is independent of a location of objects within the images.

16. The system according to claim 10, wherein a value of the threshold determines a number of images in the respective categories.

17. The system according to claim 10, wherein the database is a multimedia database.

18. The system according to claim 10, wherein the database is on the World Wide Web (WWW).