Determining a Dominant Color of an Image Based on Pixel Components

• US 20160171954A1
• Filed: 12/15/2014
• Published: 06/16/2016
• Est. Priority Date: 12/15/2014
• Status: Active Grant

• Current Assignee: Flipboard Incorporated
• Original Assignee: Flipboard Incorporated
• Inventors: Guo, Meng
• Granted Patent: US 9,576,554 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 18/22 Matching criteria, e.g. proximity measures
• CPC Class Codes: G06F 18/23 Clustering techniques
• CPC Class Codes: G06F 18/24 Classification techniques
• CPC Class Codes: G06T 2207/10024 Color image
• CPC Class Codes: G06T 7/60 Analysis of geometric attributes
• CPC Class Codes: G06T 7/90 Determination of colour characteristics
• CPC Class Codes: G06V 10/56 relating to colour
• CPC Class Codes: G09G 5/02 characterised by the way in which colour is displayed details of colour display specific for CRTs G09G1/28; specific for flat matrix panels other than liquid crystal displays G09G3/2003; specific for liquid crystal displays G09G3/3607

Abstract

A digital magazine server determines a dominant color present in an image using a clustering algorithm. Color components of each pixel in the image are identified used to generate vectors associated with each pixel. Based on the vectors associated with the pixels, clusters including one or more pixels are generated using a clustering algorithm (e.g., k-means). The digital magazine server generates a characteristic vector for each cluster based on the vectors included in the cluster and selects a set of clusters based on their characteristic vectors. A centroid identifying the dominant color of the image is determined from the characteristic vectors of clusters in the set.

Claims

1. A method comprising:

obtaining an image for presentation to a user, the image comprising a set of pixels;

identifying one or more color components associated with each pixel in the set of pixels;

generating a vector for each pixel based at least in part on the one or more color components associated with each pixel,

generating one or more clusters each including one or more pixels based at least in part on the generated vectors;

generating characteristic vectors for each cluster based at least in part on the generated vectors, a characteristic vector for a cluster based at least in part on the generated vectors of one or more pixels included in the cluster and specifying a color associated with the cluster;

selecting a set of the clusters based at least in part on the characteristic vectors;

determining a centroid based on the characteristic vectors for each cluster in the selected set of clusters, the centroid identifying a dominant color associated with the image; and

storing information associating the dominant color with the image.

2. The method of claim 1, wherein the one or more color components are identified from a set of color components associated with each pixel in the image based on a color model representing the image.

3. The method of claim 1, wherein generating the vector for each pixel based at least in part on the one or more color components associated with each pixel comprises:

identifying a pixel;

identifying one or more additional pixels within a threshold distance of the identified pixel; and

generating the vector for the identified pixel based at least in part on the one or more color components associated with the identified pixel and the one or more color components of the one or more additional pixels.

4. The method of claim 1, wherein generating the vector for the identified pixel based at least in part on the one or more color components associated with the identified pixel and the one or more color components of the one or more additional pixels comprises:

associating weights with the one or more color components associated with the identified pixel;

associating one or more additional weights with the one or more color components associated with the one or more additional pixels; and

generating the vector for the pixel based at least in part on the weights, the one or more color components associated with the identified pixels, the one or more additional weights, and the one or more additional pixels.

5. The method of claim 4, wherein an additional weight associated with one or more color components associated with an additional pixel is based at least in part on a distance between the additional pixel and the identified pixel.

6. The method of claim 5, wherein the additional weight associated with the one or more color components associated with the additional pixel is inversely proportional to the distance between the additional pixel and the identified pixel.

7. The method of claim 1, wherein generating one or more clusters each including one or more pixels based at least in part on the generated vectors comprises:

determining whether to include a pixel in a cluster based on a distance between a vector associated with the pixel and a mean value associated with the cluster.

8. The method of claim 7, wherein determining whether to include the pixel in the cluster based on a distance between the vector associated with the pixel and a mean value associated with the cluster comprises:

including the pixel in the cluster if the distance is less than a threshold value.

9. The method of claim 1, wherein a cluster includes less than a threshold number of pixels.

10. The method of claim 1, wherein a sum of magnitudes of vectors associated with pixels included in a cluster is less than a threshold value.

11. The method of claim 1, wherein selecting the set of the clusters based at least in part on the characteristic vectors comprises:

ranking the generated one or more clusters based on magnitudes of the characteristic vectors of the one or more clusters; and

selecting the set of clusters as clusters having at least a threshold position in the ranking.

12. The method of claim 1, wherein selecting the set of the clusters based at least in part on the characteristic vectors comprises:

scoring each of the generated one or more clusters based on the characteristic vectors associated with each the cluster and a number of pixels included in each cluster;

ranking the generated one or more clusters based on the scores; and

selecting the set of clusters as clusters having at least a threshold position in the ranking.

13. The method of claim 1, further comprising:

selecting an element for inclusion in the image based at least in part on the dominant color associated with the image; and

generating a modified image including the image and the selected element.

14. The method of claim 13, wherein selecting the element for inclusion in the image based at least in part on the dominant color associated with the image comprises:

associating a color with the element that contrasts with the dominant color.

15. A method comprising:

obtaining an image for presentation to a user, the image comprising a set of pixels;

identifying one or more color components associated with each pixel in the set of pixels;

generating a vector for each pixel based at least in part on the one or more color components associated each pixel,

generating one or more clusters each including one or more pixels based at least in part on the generated vectors;

generating characteristic vectors for each cluster based at least in part on the generated vectors, a characteristic vector for a cluster based at least in part on the generated vectors of one or more pixels included in the cluster and specifying a color associated with the cluster;

selecting a cluster based at least in part on a number of pixels included in each cluster;

determining a dominant color associated with the image as a color associated with a characteristic vector for the selected color and

storing information associating the dominant color with the image.

16. The method of claim 15, wherein selecting the cluster based at least in part on a number of pixels included in each cluster comprises:

selecting a cluster including a maximum number of pixels.

17. The method of claim 16, wherein generating the vector for each pixel based at least in part on the one or more color components associated with each pixel comprises:

identifying a pixel;

identifying one or more additional pixels within a threshold distance of the identified pixel; and

generating the vector for the identified pixel based at least in part on the one or more color components associated with the identified pixel and the one or more color components of the one or more additional pixels.

18. The method of claim 17, further comprising:

selecting an element for inclusion in the image based at least in part on the dominant color associated with the image; and

generating a modified image including the image and the selected element.

19. The method of claim 18, wherein selecting the element for inclusion in the image based at least in part on the dominant color associated with the image comprises:

associating a color with the element that contrasts with the dominant color.