Content based systems and methods for conducting spectrum color based image search

Patent metadata

Field	Value
Publication number	US-9990377-B1
Application number	US-201715824617-A
Country	US
Kind code	B1
Filing date	Nov 28, 2017
Priority date	Mar 14, 2013
Publication date	Jun 5, 2018
Grant date	Jun 5, 2018

Abstract

A method for dynamically displaying results of simultaneous two dimensional image search comprises: maintaining a first database comprising a set of records, each record comprising an image and an associated set of annotations describing the image, receiving an annotation criterion and a color scheme from a user, generating a target color scoring template from the color scheme, conducting a content based search by searching the first database to identify a set of pertinent images satisfying the annotation criterion, and performing the following operations for each image in the set of pertinent images: resizing a given image to a predefined size before generating an image color spectrum histogram for the given image, adding the given image to a pool of acceptable images if the generated image color spectrum histogram matches with the target color scoring template; and dynamically displaying images in the pool of acceptable images to the user.

First claim

What is claimed is:

1. A system comprising:
a memory comprising instructions; and
a processor configured to execute the instructions to:
receive from a user a search query for images from a database comprising the images and annotations describing each image, the search query comprising an annotation criterion, and a selection of a color scheme comprising a target hue bucket;
conduct a search of the database to identify a set of images responsive to the annotation criterion;
convert red-green-blue (RGB) data for each image in the set of images to luminance, chromaticity, and hue angle (LCH) data;
associate each channel of the LCH data with a distribution measure;
generate an image color spectrum histogram for each image in the set of images using the LCH data; and
when a hue value of the image color spectrum histogram for an image in the set of images corresponds to a hue value of the target hue bucket of the color scheme, display the image to the user in response to the search query.

2. The system of claim 1 , wherein the annotation criterion is included in the annotations associated with the identified set of images.

3. The system of claim 2 , wherein the user has an ability to filter images displayed in response to the search query by hue, by bright image colors, by light image colors, and by balanced image colors.

4. The system of claim 2 , wherein the image is displayed to the user in response to the search query further based on an image attribute selected by the user.

5. The system of claim 4 , wherein the user selected image attribute comprises a degree of image brightness, a degree of image balance, or a degree of image light.

6. The system of claim 1 , wherein the processor is further configured to designate a local representative color value for an image from the set.

7. The system of claim 1 , wherein a plurality of images from the set of images are displayed to the user, and the plurality of images are arranged for display based on a combination of visual attributes comprising at least one of image contrast, balance, depth of field, or lines.

8. The system of claim 1 , wherein the selected color scheme corresponds to a selection made on a slider along a color spectrum bar.

9. The system of claim 8 , wherein the color scheme is selected by moving the slider situated on the color spectrum bar in a horizontal motion.

10. The system of claim 9 , further configured to generate a weighted color scheme based on a position of the slider on the color spectrum bar.

11. A method comprising:
receiving from a user a search query for images from a database comprising the images and annotations describing each image, the search query comprising an annotation criterion, and a selection of a color scheme comprising a target hue bucket;
conducting a search of the database to identify a set of images responsive to the annotation criterion;
converting red-green-blue (RGB) data for each image in the set of images to luminance, chromaticity, and hue angle (LCH) data;
associating each channel of the LCH data with a distribution measure;
generating an image color spectrum histogram for each image in the set of images using the LCH data; and
when a hue value of the image color spectrum histogram for an image in the set of images corresponds to a hue value of the target hue bucket of the color scheme, displaying the image to the user in response to the search query.

12. The method of claim 11 , wherein the annotation criterion is included in the set of annotations associated with the identified set of images.

13. The method of claim 12 , further comprising filtering images displayed in response to the search query by hue, by bright image colors, by light image colors, or by balanced image colors selected by the user.

14. The method of claim 13 , wherein the image is displayed to the user in response to the search query further based on an image attribute selected by the user.

15. The method of claim 14 , wherein the user selected image attribute comprises a degree of image brightness, a degree of image balance, or a degree of image light.

16. The method of claim 11 , wherein a plurality of images having a hue value of their image color spectrum histogram corresponding to the hue value of the target hue bucket of the color scheme are displayed to the user in response to the search query, and the method further comprises color sorting the plurality of images at query time, and caching the sorted set of images for a subsequent use.

17. The method of claim 11 , wherein a plurality of images having a hue value of their image color spectrum histogram corresponding to the hue value of the target hue bucket of the color scheme and satisfying the annotation criterion associated with a given search term are displayed to the user in response to the search query, and the method further comprises pre-computing and caching the plurality of images in a list of popular search terms.

18. The method of claim 11 , wherein the color scheme is selected by moving the slider situated on the color spectrum bar in a horizontal motion.

19. The method of claim 18 , further comprises generating a weighted color scheme based on a position of the slider on the color spectrum bar.

20. A non-transitory machine-readable storage medium comprising machine-readable instructions for causing a processor to execute a method comprising:
receiving from a user a search query for images from a database comprising the images and annotations describing each image, the search query comprising an annotation criterion, and a selection of a color scheme comprising a target hue bucket;
conducting a search of the database to identify a set of images responsive to the annotation criterion;
converting red-green-blue (RGB) data for each image in the set of images to luminance, chromaticity, and hue angle (LCH) data;
associating each channel of the LCH data with a distribution measure;
generating an image color spectrum histogram for each image in the set of images using the LCH data; and
when a hue value of the image color spectrum histogram for an image in the set of images corresponds to a hue value of the target hue bucket of the color scheme, displaying the image to the user in response to the search query.

Assignees

• Shutterstock Inc

Inventors

• Chester David
• Jenkins Wyatt
• Rodland Andrew
• Smiling Eric
• Becker Chris

Classifications

• G06V10/56

  relating to colour · CPC title

• G06T3/40

  Scaling of whole images or parts thereof, e.g. expanding or contracting · CPC title

• G06K9/4642

  Physics · mapped topic

• G06F17/3025 Primary

  Physics · mapped topic

• G06F17/30268

  Physics · mapped topic