TWO-PASS SEARCHING FOR IMAGE SIMILARITY OF DIGESTS OF IMAGE-BASED LISTINGS IN A NETWORK-BASED PUBLICATION SYSTEM

• US 20110238659A1
• Filed: 03/28/2011
• Published: 09/29/2011
• Est. Priority Date: 03/29/2010
• Status: Active Grant

• Current Assignee: eBay Inc.
• Original Assignee: eBay Inc.
• Inventors: Henderson, Richard D., McDonald, Ryan, Roper, John, Chittar, Naren, Ghatare, Sanjay Pundlkrao
• Granted Patent: US 9,405,773 B2
• Time in Patent Office: Days
• US Class Current: 707/724
• CPC Class Codes: G06F 16/532 Query formulation, e.g. graphical querying
• CPC Class Codes: G06F 16/5838 using colour
• CPC Class Codes: G06F 16/5846 using extracted text
• CPC Class Codes: G06F 16/5854 using shape and object relationship

Abstract

Using a processor, receiving, a query including a query image. A database is searched for a set of images similar to the query image, using a two-pass search. The results of the searching are then provided, the results including image members of the set of images similar to the query image. The first pass may be performed using a TF-IDF algorithm and the second pass ranks a predetermined number of results from the first pass by a best match algorithm or other type of algorithm. The type of second pass algorithm may be selectable by a user. If an end signal is not detected, a further two-pass search may be made based on a subsequent query having a subsequent query image. The subsequent query image may one of the results of the previous two pass search. A subsequent query image may change from an image of a product of a first product category to an image of a product of a second product category.

Claims

1. A computer implemented method comprising:

receiving, using a processor, a query including a query image;

searching a database for a set of images similar to the query image, the searching of the database including using a two-pass search; and

providing results of the searching, the results including image members of the set of images similar to the query image.

2. The method of claim 1, wherein using the two-pass search comprises performing a first pass by comparing the query image against a plurality of images in the database, determining a plurality of matching images resulting from the comparing, and storing a predetermined number of matching images.

3. The method of claim 2, wherein using the two-pass search comprises performing a second pass by ranking at least some of the predetermined number of matching images using a best match procedure.

4. The method of claim 2, wherein using the two-pass search comprises performing a second pass by ranking at least some of the predetermined number of matching images based on a distance between color histograms of images having color attributes.

5. The method of claim 2, wherein using the two-pass search comprises performing a second pass by ranking at least some of the predetermined number of matching images based on cosine similarity of edge matrices of the at least some of the predetermined number of matching images having edge attributes computed using edge matrices.

6. The method of claim 2, wherein using the two-pass search comprises performing a second pass by ranking at least some of the predetermined number of matching images using a pattern comparison of the at least some of the predetermined number of matching images having pattern attributes.

7. The method of claim 2, wherein using the two-pass search comprises performing a second pass by ranking at least some of the predetermined number of matching images based on text similarity of the at least some of the predetermined number of matching images having text attributes.

8. The method of claim 2, wherein at least one of the predetermined number of the plurality of images comprises sale attributes, the second pass comprising ranking at least some of the predetermined number of matching images by at least one of the sale attributes.

9. The method of claim 1, wherein at least one of the image members when rendered is selectable as a subsequent query image.

10. The method of claim 1, further comprising

determining if an end signal is received;

and

based on the end signal not being received, performing a further two-pass search based on a subsequent query having a subsequent query image.

11. The method of claim 8, wherein the sale attributes include at least one of the group consisting of time (sale ending soonest);

newly listed;

price+shipping cost (lowest first);

price+shipping cost (highest first);

price (highest first);

price (lowest first), seller and item quality.

12. The method of claim 10 wherein a subsequent query image changes from an image of a product of a first product category to an image of a product of a second product category.

13. The method of claim 1 wherein the query includes structured and unstructured data.

14. The method of claim 1 wherein the second pass is selectable by a user.

15. The method of claim 2 wherein the first pass is made using a TF-IDF algorithm.

16. A non-transitory computer-readable storage medium having embedded therein a set of instructions which, when executed by one or more processors of a computer, causes the computer to execute the following operations:

receiving a query including a query image;

searching a database for a set of images similar to the query image, the searching of the database including using a two-pass search; and

providing results of the searching, the results including image members of the set of images similar to the query image.

17. The medium of claim 16, wherein using the two-pass search comprises performing a first pass by comparing the query image against a plurality of images in the database, determining a plurality of matching images resulting from the comparing, and storing a predetermined number of matching images.

18. The medium of claim 16, wherein using the two-pass search comprises performing a second pass by ranking at least some of the predetermined number of matching images using a best match procedure.

19. The medium of claim 16, wherein using the two-pass search comprises performing a second pass by ranking at least some of the predetermined number of matching images based on a distance between color histograms of images having color attributes.

20. The medium of claim 16, wherein using the two-pass search comprises performing a second pass by ranking at least some of the predetermined number of matching images based on cosine similarity of edge matrices of the at least some of the predetermined number of matching images having edge attributes computed using edge matrices.

21. The medium of claim 16, wherein using the two-pass search comprises performing a second pass by ranking at least some of the predetermined number of matching images using a pattern comparison of the at least some of the predetermined number of matching images having pattern attributes.

22. The medium of claim 16, wherein using the two-pass search comprises performing a second pass by ranking at least some of the predetermined number of matching images based on text similarity of the at least some of the predetermined number of matching images having text attributes.

23. The medium of claim 16, wherein at least one of the predetermined number of the plurality of images comprises sale attributes, the second pass comprising ranking at least some of the predetermined number of matching images by at least one of the sale attributes.

24. The medium of claim 16, wherein at least one of the image members when rendered is selectable as a subsequent query image.

25. The medium of claim 16, further comprising

determining if an end signal is received;

and

based on the end signal not being received, performing a further two-pass search based on a subsequent query having a subsequent query image.

26. The medium of claim 23, wherein the sale attributes include at least one of the group consisting of time (sale ending soonest);

newly listed;

price+shipping cost (lowest first);

price+shipping cost (highest first);

price (highest first);

price (lowest first), seller and item quality.

27. The method of claim 25 wherein a subsequent query image changes from an image of a product of a first product category to an image of a product of a second product category

28. The method of claim 16 wherein the query includes structured and unstructured data.

29. The medium of claim 16 wherein the first pass is made using a TF-IDF algorithm.

30. A system comprising:

a computer processor and storage configured to store and execute

a receiving module for receiving a query including a query image;

a plurality of search algorithms for searching a database for a set of images similar to the query image, the searching of the database including using a two-pass search, one of the search algorithms being a TF-IDF algorithm; and

a transmission circuit for transmitting the results of searching the database, the results including image members of the set of images similar to the query image.