Methods and apparatus to perform image classification based on pseudorandom features

• US 8,818,112 B2
• Filed: 12/07/2012
• Issued: 08/26/2014
• Est. Priority Date: 04/27/2009
• Status: Active Grant

• Current Assignee: The Nielsen Company LLC (MOOD Media Corporation)
• Original Assignee: The Nielsen Company LLC (MOOD Media Corporation)
• Inventors: Lee, Morris
• Primary Examiner(s): STREGE, JOHN B
• Application Number:
  US13/708,660
• Publication Number:
  US 20130108153A1
• Time in Patent Office: 627 Days
• Field of Search: 382/224, 382/159, 382/118
• US Class Current: 382/224
• CPC Class Codes: G06F 18/21 Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
• CPC Class Codes: G06F 18/241 relating to the classification model, e.g. parametric or non-parametric approaches
• CPC Class Codes: G06V 10/40 Extraction of image or video features
• CPC Class Codes: G06V 10/70 using pattern recognition or machine learning optical pattern recognition or electronic computations therefor G06V10/88
• CPC Class Codes: G06V 10/764 using classification, e.g. of video objects
• CPC Class Codes: G06V 40/16 Human faces, e.g. facial parts, sketches or expressions

Abstract

Example methods and apparatus to perform image classification based on pseudorandom features are disclosed. A disclosed example method includes generating first and second pseudorandom numbers, extracting a first feature of an image based on the first and second pseudorandom numbers, and determining a classification for the image based on the first extracted feature.

Claims

1. A method comprising:

generating first and second pseudorandom numbers;

extracting a first feature of an image based on the first and second pseudorandom numbers, wherein extracting the first feature of the image comprises;

identifying a location of a region of the image based on the first and second pseudorandom numbers; and

computing an average of pixels of the identified region;

determining whether a classification for the image is known; and

in response to a known classification of the image, updating a classification model used to classify a second image.

2. A method as defined in claim 1, wherein generating the first pseudorandom number further comprises:

resetting a pseudorandom number generation seed;

generating a pseudorandom number sequence using the pseudorandom number generation seed; and

selecting the first pseudorandom number from the pseudorandom number sequence.

3. A method as defined in claim 1, further comprising:

in response to an unknown classification of the image;

forming an image feature vector based on the first feature;

scaling the image feature vector with one or more classification parameters; and

classifying the image based on the scaled image feature vector.

4. A method as defined in claim 3, wherein classifying the image further comprises:

providing the first feature into a support vector machine, wherein the support vector machine determines in which identified location the first feature is located; and

classifying the image based on the support vector machine determination.

5. A method as defined in claim 1, wherein the classification comprises a gender of a person.

6. A method as defined in claim 1, wherein extracting the first feature further comprises:

computing at least one of a distance between eyes, a dimension of an area defined by eyes and a nose, or a mouth dimension.

7. A method as defined in claim 1, wherein updating the classification model further comprises updating one or more image vector scale factors.

8. A method as defined in claim 1, wherein updating the classification model further comprises updating one or more classification hyperplanes.

9. An apparatus comprising:

a pseudorandom number generator to generate one or more pseudorandom numbers;

a feature extractor to extract a first feature of an image based on the one or more pseudorandom numbers by identifying a region of the image based on the one or more pseudorandom numbers and computing an average of pixels of the identified region;

an image classifier to determine whether a classification for the image is known, and in response to a known classification of the image, updating one or more classification parameters, wherein the one or more classification parameters are used to classify a second image.

10. An apparatus as defined in claim 9, wherein the pseudorandom number generator is to:

reset a pseudorandom number generation seed;

generate a pseudorandom number sequence using the pseudorandom number generation seed; and

select the first pseudorandom number from the pseudorandom number sequence.

11. An apparatus as defined in claim 9, wherein the feature extractor is to compute at least one of a distance between eyes, a dimension of an area defined by eyes and a nose, or a mouth dimension.

12. An apparatus as defined in claim 9, wherein the image classifier is to update one or more image vector scale factors.

13. An apparatus as defined in claim 9, wherein, in response to an unknown classification of the image, the image classifier is to:

form an image feature vector based on the first feature;

scale the image feature vector with one or more classification parameters; and

classify the image based on the scaled image feature vector.

14. An apparatus as defined in claim 13, wherein the image classifier is to:

determine in which identified location the first feature is located; and

classify the image based on the determination.

15. A tangible computer readable storage device or storage disc comprising machine-readable instructions which, when executed, cause a machine to at least:

generate one or more pseudorandom numbers;

extract a first feature of an image based on the one or more pseudorandom numbers by identifying a region of the image based on the one or more pseudorandom numbers and computing an average of pixels of the identified region;

determine whether a classification for the image is known; and

in response to a known classification of the image, update one or more classification parameters, wherein the one or more classification parameters are used to classify a second image.

16. A tangible computer readable storage device or storage disc as defined in claim 15, wherein the instructions further cause the machine to:

reset a pseudorandom number generation seed;

generate a pseudorandom number sequence using the pseudorandom number generation seed; and

select the first pseudorandom number from the pseudorandom number sequence.

17. A tangible computer readable storage device or storage disc as defined in claim 15, wherein the instructions further cause the machine to:

to compute at least one of a distance between eyes, a dimension of an area defined by eyes and a nose, or a mouth dimension.

18. A tangible computer readable storage device or storage disc as defined in claim 15, wherein the instructions further cause the machine to:

to update one or more image vector scale factors.

19. A tangible computer readable storage device or storage disc as defined in claim 15, wherein the instructions further cause the machine to:

form an image feature vector based on the first feature;

scale the image feature vector with one or more classification parameters; and

classify the image based on the scaled image feature vector.

20. A tangible computer readable storage device or storage disc as defined in claim 19, wherein the instructions further cause the machine to:

determine in which identified location the first feature is located; and

classify the image based on the determination.