Addition of new images to an image database by clustering according to date/time and image content and representative image comparison

• US 7,831,599 B2
• Filed: 03/04/2005
• Issued: 11/09/2010
• Est. Priority Date: 03/04/2005
• Status: Expired due to Fees

• Current Assignee: Monument Peak Ventures, LLC (Dominion Harbor Enterprises, LLC)
• Original Assignee: Eastman Kodak Company
• Inventors: Das, Madirakshi, Loui, Alexander C.
• Primary Examiner(s): Pham; Hung Q
• Assistant Examiner(s): CHEUNG, HUBERT G
• Application Number:
  US11/072,792
• Publication Number:
  US 20060200475A1
• Time in Patent Office: 2,076 Days
• Field of Search: 707/737, 382/225
• US Class Current: 707/737
• CPC Class Codes: G06F 16/58 Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
• CPC Class Codes: G06F 16/5838 using colour

Abstract

A database has chronologically ordered images classified into event groups based upon a time difference threshold, and into subgroups based upon a similarity measure. In a method and system for combining new images into such a database, new images are ordered into clusters based upon assessed image features. A representative image is selected in each cluster. A database segment chronologically overlapping the new images is designated and a set of database images similar to each representative image are identified in the segment. Different subgroups including one or more retrieved images are associated with each cluster to provide matched subgroups. The new images are assigned to matched subgroups associated with respective clusters.

Claims

1. A method for combining a plurality of new images into a preexisting database, the method using a processor to perform the steps of:

receiving the plurality of new images;

designating the preexisting database, wherein the designated preexisting database stores chronologically ordered images, wherein the stored chronologically ordered images in the designated preexisting database were previously classified into database event groups based upon a time difference threshold, and wherein images in the database event groups were previously classified into database event subgroups based upon an image similarity measure;

ordering the received plurality of new images into chronological order to form an ordered sequence of new images;

forming new image clusters by analyzing the formed ordered sequence of new images using the image similarity measure to determine image cluster boundaries between consecutive images in the formed ordered sequence of new images;

selecting a representative image in each of the formed new image clusters;

designating a segment of the designated preexisting database, wherein the stored chronologically ordered images in the designated preexisting database chronologically overlaps the received plurality of new images; and

assigning each of the formed new image clusters to a corresponding database event group and a corresponding database event subgroup within the designated segment of the designated preexisting database by optimizing a joint match score constrained by the chronological ordering of the received plurality of new images being maintained, wherein the joint match score is responsive to image similarity measures calculated between the selected representative image in each of the formed new image clusters and images in the database event subgroups.

2. The method of claim 1, wherein the ordering of the received plurality of new images includes determining the chronological order based upon metadata associated with the received plurality of new images.

3. The method of claim 2, wherein the metadata associated with the received plurality of new images includes automatically-generated dates of file origination.

4. The method of claim 2, wherein the received plurality of new images have unique filenames inclusive of metadata defining the chronological order and wherein the ordering of the received plurality of new images includes determining the chronological order based upon the unique filenames.

5. The method of claim 1, wherein the ordering of the received plurality of new images includes determining the chronological order based upon a physical order that the received plurality of new images are stored in a storage media.

6. The method of claim 1, wherein the ordering of the received plurality of new images includes accepting user input designating the chronological order.

7. The method of claim 1, wherein the preexisting database defines a most recent database image date-time and the designating of the segment of the designated preexisting database further includes determining a last date-time of the received plurality of new images, wherein the determined last date-time being no later than the defined most recent database image date-time.

8. The method of claim 7, wherein the designating of the segment of the designated preexisting database further includes determining a beginning date-time of the received plurality of new images, and limiting the designated segment of the designated preexisting database to the stored chronologically ordered images on or between the determined beginning date-time and the determined last date-time.

9. The method of claim 8, wherein the determined beginning date-time is determined by accepting user input designating the determined beginning date-time.

10. The method of claim 1, wherein the optimizing the joint match score includes minimizing a cumulative temporal duration of matched database event subgroups.

11. The method of claim 10, wherein a greedy algorithm is used to minimize the cumulative temporal duration of the matched database event subgroups.

12. The method of claim 1, wherein the optimizing of the joint match score includes maximizing the image similarity measures calculated between the selected representative image in each of the formed new image clusters and the images in the database event subgroups.

13. The method of claim 1, wherein the optimizing of the joint match score further comprises:

assessing a variance of the image similarity measures calculated between the selected representative image in each of the formed new clusters and matched images from the database event subgroups;

when the assessed variance is less than a predetermined threshold, assigning the database event subgroups to minimize a cumulative temporal duration of the database event subgroups; and

when the assessed variance is not less than the predetermined threshold, assigning the database event subgroups to maximize the image similarity measures calculated between the selected representative image in each of the formed new image clusters and the images in the database event subgroups.

14. The method of claim 1, wherein the designated segment of the designated preexisting database includes all of the stored chronologically ordered images chronologically earlier than a last date-time of the received plurality of new images.

15. A system for combining a plurality of new images into a preexisting database, the system comprising:

computer readable storage media containing;

the preexisting database, wherein the preexisting database stores chronologically ordered images, wherein the stored chronologically ordered images in the preexisting database were previously classified into database event groups based upon a time difference threshold, and wherein images in the database event groups were previously classified into database event subgroups based upon an image similarity measure; and

a plurality of new images; and

a microprocessor unit communicably connected to the computer readable storage media and configured to;

receive the plurality of new images;

designate the preexisting database;

order the received plurality of new images into chronological order to form an ordered sequence of new images;

form new image clusters by analyzing the formed ordered sequence of new images using the image similarity measure to determine image cluster boundaries between consecutive images in the formed ordered sequence of new images;

select a representative image in each of the formed new image clusters;

designate a segment of the designated preexisting database, wherein the stored chronologically ordered images in the designated preexisting database chronologically overlaps the received plurality of new images; and

assign each of the formed new image clusters to a corresponding database event group and a corresponding database event subgroup within the designated segment of the designated preexisting database by optimizing a joint match score constrained by the chronological ordering of the received plurality of new images being maintained, wherein the joint match score is responsive to image similarity measures calculated between the selected representative image in each of the formed new image clusters and images in the database event subgroups.

16. A computer program product for combining a plurality of new images into a preexisting database, the computer program product comprising a computer readable storage medium storing a computer program executable by a computer for executing a method comprising the steps of:

receiving the plurality of new images;

designating the preexisting database, wherein the designated preexisting database stores chronologically ordered images, wherein the stored chronologically ordered images in the designated preexisting database were previously classified into database event groups based upon a time difference threshold, and wherein images in the database event groups were previously classified into database event subgroups based upon an image similarity measure;

ordering the received plurality of new images into chronological order to form an ordered sequence of new images;

forming new image clusters by analyzing the formed ordered sequence of new images using the image similarity measure to determine image cluster boundaries between consecutive images in the formed ordered sequence of new images;

selecting a representative image in each of the formed new image clusters;

designating a segment of the designated preexisting database, wherein the stored chronologically ordered images in the designated preexisting database chronologically overlaps the received plurality of new images; and

assigning each of the formed new image clusters to a corresponding database event group and a corresponding database event subgroup within the designated segment of the designated preexisting database by optimizing a joint match score constrained by the chronological ordering of the received plurality of new images being maintained, wherein the joint match score is responsive to image similarity measures calculated between the selected representative image in each of the formed new image clusters and images in the database event subgroups.