Method, apparatus, and computer program product for generating an image index and for internet searching and querying by image colors

• US 6,026,411 A
• Filed: 11/06/1997
• Issued: 02/15/2000
• Est. Priority Date: 11/06/1997
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Delp, Helen Roxlo
• Primary Examiner(s): Black, Thomas
• Assistant Examiner(s): Coby, Frantz
• Application Number:
  US08/965,621
• Time in Patent Office: 831 Days
• Field of Search: 707/1, 707/3, 707/101, 707/104, 382/162, 382/169
• US Class Current: 707/700
• CPC Class Codes: G06F 16/5838 using colour
• CPC Class Codes: Y10S 707/915 Image
• CPC Class Codes: Y10S 707/99933 Query processing, i.e. searching
• CPC Class Codes: Y10S 707/99945 Object-oriented database structure processing

Abstract

A method, apparatus and computer program product are provided for building an image index and for searching and querying by image colors from a plurality of images including images from the internet or other network environment. First an index is built for each of the plurality of images. A selected number of levels is identified for image processing, and an upper resolution and a lower resolution are identified for each of the selected number of levels for image processing. An image is filtered at each selected level. A sample of pixels is acquired and stored at each selected level in the index. At query time, an image color is selected. A entry of the index is acquired. The acquired entry is compared with the selected image color. A match is identified responsive to the acquired entry being near the selected image color.

Claims

1. A computer program product for use in a computer system for building an index by image colors from a plurality of images including images from the internet, said computer program product comprising:

a recording medium;

means, recorded on the recording medium, for identifying a selected number of levels for image processing, and for identifying an upper resolution and a lower resolution for each of said selected number of levels for image processing;

means, recorded on the recording medium, for filtering an image at each said selected level for image processing of said filtered image; and

means, recorded on the recording medium, for acquiring and storing a sample of pixels at each said selected level for image processing of said filtered image.

2. A computer program product as recited in claim 1 includes means, recorded on the recording medium, for identifying a selected number of important colors.

3. A computer program product as recited in claim 1 wherein said means, recorded on the recording medium, for acquiring and storing said sample of pixels at each selected level includes means, recorded on the recording medium, for utilizing a random, a pseudo-random or a deterministic method, and for acquiring and storing said sample of pixels at each said selected level for image processing of said filtered image in said index.

4. A computer program product as recited in claim 1 wherein said means, recorded on the recording medium, for acquiring and storing said sample of pixels at each selected level includes means, recorded on the recording medium, for generating a histogram of pixels by image color and for utilizing said histogram for acquiring and storing said sample of pixels at each said selected level for image processing of said filtered image in said index.

5. A computer program product as recited in claim 4 wherein said means, recorded on the recording medium, for acquiring and storing said sample of pixels at each said selected level includes means, recorded on the recording medium, for identifying a selected number of important colors and for utilizing said selected number of important colors and said histogram for acquiring and storing said sample of pixels at each said selected level for image processing of said filtered image in said index.

6. A computer program product as recited in claim 4 includes means, recorded on the recording medium, for identifying a histogram threshold for selecting a number of important colors.

7. A computer program product for use in a computer system for querying by image colors from a plurality of images including images from the internet, said computer program product comprising:

a recording medium;

an index by image colors from the plurality of images including images from the internet stored on said recording medium;

means, recorded on the recording medium, for selecting an image color;

means, recorded on the recording medium, for acquiring an index entry from said index;

means, recorded on the recording medium, for comparing said acquired index entry with said selected image color; and

means, recorded on the recording medium, for identifying a match responsive to said acquired index entry being near said selected image color.

8. A computer program product as recited in claim 7 includes means, recorded on the recording medium, for building said index for each of the plurality of images.

9. A computer program product as recited in claim 7 includes said means, recorded on the recording medium, for identifying a selected coordinate system for comparing said acquired index entry with said selected image color.

10. A computer program product as recited in claim 7 includes means, recorded on the recording medium, for calculating a match quality value responsive to each identified match of said acquired index entry and said selected image color.

11. A computer program product as recited in claim 7 wherein said means, recorded on the recording medium, for identifying said match responsive to said acquired index entry being near said selected image color includes means, recorded on the recording medium, for identifying a permissible color difference between said acquired pixel and said selected image color for identifying a match.

12. A computer program product as recited in claim 7 wherein said means, recorded on the recording medium, for identifying said match responsive to said acquired entry being near said selected image color includes means, recorded on the recording medium, for identifying a predefined distance or for identifying a predefined range, and wherein said means, recorded on the recording medium, for identifying said match is responsive to said acquired index entry being within said predefined distance or within said predefined range of said selected image color.

13. A computer program product as recited in claim 7 wherein said means, recorded on the recording medium, for identifying said match responsive to said acquired entry being near said selected image color includes means, recorded on the recording medium, for identifying said match is responsive to said acquired index entry being equal to said selected image color.

14. A computer implemented method for building an index by image colors from a plurality of images including images from the internet, said method comprising the steps of:

identifying a selected number of levels for image processing, and identifying an upper resolution and a lower resolution for each of said selected number of levels for image processing;

filtering an image at each selected level between said upper resolution and said lower resolution for image processing; and

acquiring and storing a sample of pixels at each selected level from said filtered image.

15. A method for querying by image colors from a plurality of images as recited in claim 14 wherein the steps of identifying said selected number of levels for image processing, and identifying said upper resolution and said lower resolution includes the step of either receiving user selections or identifying stored default values.

16. A computer implemented method for building an index by image colors as recited in claim 14 includes the step of identifying a selected number of important colors.

17. A computer implemented method for building an index by image colors as recited in claim 14 wherein the step of acquiring and storing said sample of pixels at each selected level from said filtered image includes the steps of, utilizing a random, a pseudo-random or a deterministic method, and acquiring and storing a sample of pixels at each selected level in said index.

18. A computer implemented method for building an index by image colors as recited in claim 14 wherein the step of acquiring and storing said sample of pixels at each selected level from said filtered image includes the steps of generating a histogram of pixels by image color and utilizing said histogram for acquiring and storing a sample of pixels at each selected level in said index.

19. A computer implemented method for building an index by image colors as recited in claim 18 wherein the step of acquiring and storing said sample of pixels at each selected level from said filtered image includes the steps of identifying a selected number of important colors and utilizing said selected number of important colors and said histogram for acquiring and storing said sample of pixels at each selected level in said index.

20. A computer implemented method for building an index by image colors as recited in claim 18 includes the step of identifying a histogram threshold for selecting a number of important colors.

21. A method for querying by image colors from a plurality of images including images from the internet comprising the steps of:

building an index by image colors from a plurality of images including images from the internet;

selecting an image color;

acquiring an index entry from said index;

comparing said acquired index entry with said selected image color; and

identifying a match responsive to said acquired index entry being near said selected image color.

22. A method for querying by image colors from a plurality of images as recited in claim 21 includes the step of calculating a match quality value responsive to each identified match of said acquired index entry and said selected image color.

23. A method for querying by image colors from a plurality of images as recited in claim 21 wherein the step of identifying said match responsive to said acquired index entry being near said selected image color includes the step of identifying a selected coordinate system for comparing said acquired index entry with said selected image color.

24. A method for querying by image colors from a plurality of images as recited in claim 21 wherein the steps of comparing said acquired index entry with said selected image color;

and identifying said match responsive to said acquired index entry being near said selected image color include the step of identifying said match responsive to said acquired index entry being within a predefined distance of said selected image color.

25. A method for querying by image colors from a plurality of images as recited in claim 21 wherein the steps of comparing said acquired index entry with said selected image color;

and identifying said match responsive to said acquired index entry being near said selected image color include the step of identifying said match responsive to said acquired entry being within a defined range of said selected image color.

26. A method for querying by image colors from a plurality of images as recited in claim 21 wherein the steps of comparing said acquired index entry with said selected image color;

and identifying said match responsive to said acquired index entry being near said selected image color include the step of identifying said match responsive to said acquired index entry being equal to said selected image color.

27. A computer system having apparatus for building an index by image colors for a plurality of images including images from the internet, said computer system comprising:

a processor,

a memory;

a communications adapter;

a bus connecting said processor, memory, and communications adapter;

means for identifying a selected number of levels for image processing, and for identifying an upper resolution and a lower resolution for each of said selected number of levels for image processing;

means for filtering an image at each selected level for image processing between said upper resolution and said lower resolution; and

means for acquiring and for storing a sample of pixels at each selected level from said filtered image.

28. A computer system having apparatus for querying by image colors from a plurality of images including images from the internet, said computer system comprising:

a processor,

a memory;

a communications adapter;

a bus connecting said processor, memory, and communications adapter;

means for building an index by image colors from the plurality of images;

means for selecting an image color;

means for acquiring an index entry from said index;

means for comparing said acquired index entry with said selected image color; and

means for identifying a match responsive to said acquired index entry being near said selected image color.