Detecting objects in images using color histograms

• US 8,750,613 B2
• Filed: 12/13/2011
• Issued: 06/10/2014
• Est. Priority Date: 12/13/2011
• Status: Active Grant

• Current Assignee: Nielsen Company (US) LLC (RR Donnelley & Sons Company)
• Original Assignee: The Nielsen Company LLC (MOOD Media Corporation)
• Inventors: Lee, Morris
• Primary Examiner(s): Tran, Phuoc
• Application Number:
  US13/324,695
• Publication Number:
  US 20130148882A1
• Time in Patent Office: 910 Days
• Field of Search: 382/103, 382/164, 382/168, 382/171, 382/173, 382/170
• US Class Current: 382/168
• CPC Class Codes: G06V 10/255 Detecting or recognising potential candidate objects based on visual cues, e.g. shapes
• CPC Class Codes: G06V 10/50 by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis
• CPC Class Codes: G06V 10/56 relating to colour

Abstract

Methods, apparatus and articles of manufacture for detecting objects in images using color histograms are disclosed. An example method disclosed herein comprises segmenting an image into sets of subregions, different sets having at least one of different subregion sizes or different subregion offsets relative to one another, determining a first comparison metric quantifying comparison of a first subregion with the object based on differences between bin values of a second color histogram corresponding to the object and adjusted bin values of a first color histogram corresponding to the first subregion, determining a second comparison metric quantifying comparison of a second subregion with the object, and comparing the first and second comparison metrics with a threshold to identify at least one of the subregions as being associated with a possible location of the object in the image.

Claims

1. A method to detect an object in an image, the method comprising:

segmenting the image into sets of subregions, different sets having at least one of different subregion sizes or different subregion offsets relative to one another;

determining a first comparison metric quantifying comparison of a first subregion with the object based on differences between bin values of a second color histogram corresponding to the object and adjusted bin values of a first color histogram corresponding to the first subregion;

determining a second comparison metric quantifying comparison of a second subregion with the object; and

comparing the first and second comparison metrics with a threshold to identify at least one of the subregions as being associated with a possible location of the object in the image.

2. A method as defined in claim 1 further comprising:

scaling the bin values of the first color histogram by a scale factor to determine scaled bin values of the first color histogram; and

determining an adjusted bin value of the first color histogram for a particular color bin to be a smaller of a scaled bin value of the first color histogram for the particular color bin and a respective bin value of the second color histogram for the particular color bin.

3. A method as defined in claim 2 wherein the scale factor is substantially equal to two.

4. A method as defined in claim 1 wherein determining the first comparison metric comprises:

determining the differences between the bin values of the second color histogram and the respective adjusted bin values of the first color histogram; and

summing the differences to determine the first comparison metric.

5. A method as defined in claim 1 further comprising obtaining the first color histogram corresponding to the first subregion by:

quantizing color values of pixels forming the first subregion; and

including the pixels in color bins of the first color histogram, a first pixel being included in a first one of the color bins based on a combination of quantized color values of the first pixel and quantized color values of a set of one or more other pixels in a neighborhood of the first pixel.

6. A method as defined in claim 1 further comprising:

identifying a subset of the sets of subregions having respective comparison metrics that meet the threshold; and

depicting the subset of the sets of subregions in a presentation of the image to depict possible locations of the object in the image.

7. A method as defined in claim 1 further comprising:

identifying a subset of the sets of subregions having respective comparison metrics that meet the threshold; and

combining location information for the subregions in the subset to estimate a location of the object in the image.

8. A tangible machine readable medium storing machine readable instructions which, when executed, cause a machine to at least:

segment an image into sets of subregions, different sets having at least one of different subregion sizes or different subregion offsets relative to one another;

determine a first comparison metric quantifying comparison of a first subregion with the object based on differences between bin values of a second color histogram corresponding to the object and adjusted bin values of a first color histogram corresponding to the first subregion;

determine a second comparison metric quantifying comparison of a second subregion with the object; and

compare the first and second comparison metrics with a threshold to identify at least one of the subregions as being associated with a possible location of the object in the image.

9. A tangible machine readable medium as defined in claim 8 wherein the machine readable instructions, when executed, further cause the machine to:

scale the bin values of the first color histogram by a scale factor to determine scaled bin values of the first color histogram; and

determine an adjusted bin value of the first color histogram for a particular color bin to be a smaller of a scaled bin value of the first color histogram for the particular color bin and a respective bin value of the second color histogram for the particular color bin.

10. A tangible machine readable medium as defined in claim 9 wherein the scale factor is substantially equal to two.

11. A tangible machine readable medium as defined in claim 8 wherein the machine readable instructions, when executed, further cause the machine to:

determine the differences between the bin values of the second color histogram and the respective adjusted bin values of the first color histogram; and

sum the differences to determine the first comparison metric.

12. A tangible machine readable medium as defined in claim 8 wherein to obtain the first color histogram corresponding to the first subregion, the machine readable instructions, when executed, further cause the machine to:

quantize color values of pixels forming the first subregion; and

include the pixels in color bins of the first color histogram, a first pixel being included in a first one of the color bins based on a combination of quantized color values of the first pixel and quantized color values of a set of one or more other pixels in a neighborhood of the first pixel.

13. A tangible machine readable medium as defined in claim 8 wherein the machine readable instructions, when executed, further cause the machine to:

identify a subset of the sets of subregions having respective comparison metrics that meet the threshold; and

depict the subset of the sets of subregions in a presentation of the image to depict possible locations of the object in the image.

14. A tangible machine readable medium as defined in claim 8 wherein the machine readable instructions, when executed, further cause the machine to:

identify a subset of the sets of subregions having respective comparison metrics that meet the threshold; and

combine location information for the subregions in the subset to estimate a location of the object in the image.

15. An apparatus to detect an object in an image, the apparatus comprising:

a histogram generator to generate respective test color histograms for subregions in sets of subregions of the image, different sets having at least one of different subregion sizes or different subregion offsets relative to one another; and

a histogram comparator to;

determine comparison metrics quantifying comparison of respective ones of the subregions with the object, a first comparison metric quantifying comparison of a first subregion with the object being based on differences between bin values of a reference color histogram corresponding to the object and respective adjusted bin values of a respective one of the test color histograms corresponding to the first subregion; and

compare the comparison metrics with a threshold to identify at least one of the subregions as being associated with a possible location of the object in the image.

16. An apparatus as defined in claim 15 wherein the histogram comparator is to:

scale bin values of the test color histogram by a scale factor to determine the scaled bin values of the test color histogram; and

determine an adjusted bin value of the test color histogram for a particular color bin to be a smaller of a scaled bin value of the test color histogram for the particular color bin and a respective bin value of the reference color histogram for the particular color bin.

17. An apparatus as defined in claim 15 wherein the scale factor is substantially equal to two.

18. An apparatus as defined in claim 15 wherein the histogram comparator is to determine the first comparison metric to be a sum of differences between the bin values of the reference color histogram and the respective adjusted bin values of the test color histogram.

19. An apparatus as defined in claim 15 wherein to generate the test color histogram corresponding to the first subregion, the histogram generator is to:

quantize color values of pixels forming the first subregion; and

include the pixels in color bins of the test color histogram, a first pixel being included in a first one of the color bins based on a combination of quantized color values of the first pixel and quantized color values of a set of one or more other pixels in a neighborhood of the first pixel.

20. An apparatus as defined in claim 15 further comprising an object locator to:

identify a subset of the sets of subregions having respective comparison metrics that meet the threshold;

depict the subset of the sets of subregions in a presentation of the image to depict possible locations of the object in the image; and

combine location information for the subregions in the subset to estimate a location of the object in the image.