Method and apparatus for providing image classification based on opacity

• US 9,870,511 B2
• Filed: 10/14/2015
• Issued: 01/16/2018
• Est. Priority Date: 10/14/2015
• Status: Active Grant

• Current Assignee: HERE Global B.V. (Nokia Corporation)
• Original Assignee: HERE Global B.V. (Nokia Corporation)
• Inventors: Mittal, Anish, Marks, William
• Primary Examiner(s): Kassa, Yosef
• Application Number:
  US14/883,274
• Publication Number:
  US 20170109612A1
• Time in Patent Office: 825 Days
• Field of Search: 382165, 382170, 382171, 382190, 382209, 382224, 382264, 382274, 382275, 358 326, 358 327, 358537
• CPC Class Codes: G06V 20/56 exterior to a vehicle by using sensors mounted on the vehicle

Abstract

An approach is provided for automated classification of an image based on the fogging attributes associated with the image. The approach involves processing and/or facilitating a processing of image data associated with at least one image to cause, at least in part, an extraction of one or more fogging attributes from the image data. The approach also involves causing, at least in part, a mapping of the one or more fogging attributes to one or more features of at least one classifier, wherein the at least one classifier is trained based, at least in part, on a co-registration of the one or more features to one or more records of previously made image judgements. The approach further involves causing, at least in part, a classification of the at least one image as either in a fogging-afflicted state or a non-fogging-afflicted state based, at least in part, on the at least one classifier.

Claims

1. A method comprising:

determining, by an apparatus, image data from a plurality of bounding boxes each centering at an identical point of an image, wherein the bounding boxes are of different sizes each smaller than a size of the image;

initiating, by the apparatus, an extraction of one or more fogging attributes from the image data;

initiating, by the apparatus, a mapping of the one or more fogging attributes to one or more features of at least one classifier, wherein the at least one classifier is trained based, at least in part, on a co-registration of the one or more features to one or more records of previously made image judgements; and

initiating, by the apparatus, a classification of the image as either in a fogging-afflicted state or a non-fogging-afflicted state based, at least in part, on the at least one classifier.

2. A method of claim 1,

wherein the bounding boxes are located within a center region of the image.

3. A method of claim 2, wherein the identical point of the image is the center point of the image, the bounding boxes include, at least in part, one or more small bounding boxes, one or more medium bounding boxes, and one or more large bounding boxes specified for a plurality of color channels of the image.

4. A method of claim 3, further comprising:

determining at least one multi-dimensional contrast attribute vector based, at least in part, on the bounding boxes for every pixel in each of the color channels.

5. A method of claim 3, further comprising:

computing an average value of contrast of pixels per bounding box for each of the one or more small bounding boxes, the one or more medium bounding boxes, and the one or more large bounding boxes in each of the color channels; and

including the average value of contrast in the one or more fogging attributes.

6. A method of claim 1, wherein the one or more features of the at least one classifier include, at least in part, one or more color-based localized contrast features.

7. A method of claim 6, further comprising:

determining the one or more color-based localized contrast features based, at least in part, on a Michelson contrast value.

8. A method of claim 7, further comprising:

determining that the one or more color-based localized contrast features are associated with the fogging-afflicted state if the Michelson contrast value is below a threshold value; and

determining that the one or more color-based localized contrast features are associated with the non-fogging-afflicted state if the Michelson contrast value is above a threshold value.

9. A method of claim 7, further comprising:

determining the Michelson contrast value for at least one pixel of the image based, at least in part, on intensity information associated with one or more neighboring pixels.

10. A method of claim 1, further comprising:

using one or more classifying algorithms to generate the at least one classifier based, at least in part, on comparing the one or more fogging attributes to one or more subjective fogging level judgements provided by human users,

wherein the fogging-afflicted state is associated with at least a fogging or a hazing at least one image, wherein the fogging or the hazing results from one or more conditions of at least one camera component used to capture the image, one or more external conditions associated with at least one environment captured in the image, or a combination thereof.

11. An apparatus comprising:

at least one processor; and

at least one memory including computer program code for one or more programs,

the at least one memory and the computer program code configured to, with the at least one processor, cause the apparatus to perform at least the following,

determine image data from a plurality of bounding boxes each centering at an identical point of an image, wherein the bounding boxes are of different sizes each smaller than a size of the image;

initiate an extraction of one or more fogging attributes from the image data;

initiate a mapping of the one or more fogging attributes to one or more features of at least one classifier, wherein the at least one classifier is trained based, at least in part, on a co-registration of the one or more features to one or more records of previously made image judgements; and

initiate a classification of the image as either in a fogging-afflicted state or a non-fogging-afflicted state based, at least in part, on the at least one classifier.

12. An apparatus of claim 11,

wherein the bounding boxes are located within a center region of the image.

13. An apparatus of claim 12, wherein the identical point of the image is the center point of the image, the bounding boxes include, at least in part, one or more small bounding boxes, one or more medium bounding boxes, and one or more large bounding boxes specified for a plurality of color channels of the image.

14. An apparatus of claim 13, wherein the apparatus is further caused to:

determine at least one multi-dimensional contrast attribute vector based, at least in part, on the bounding boxes for every pixel in each of the color channels.

15. An apparatus of claim 12, wherein the apparatus is further caused to:

compute an average value of contrast of pixels per bounding box for each of the one or more small bounding boxes, the one or more medium bounding boxes, and the one or more large bounding boxes in each of the color channels; and

include the average value of contrast in the one or more fogging attributes.

16. An apparatus of claim 11, wherein the one or more features of the at least one classifier include, at least in part, one or more color-based localized contrast features.

17. An apparatus of claim 16, wherein the apparatus is further caused to:

determine the one or more color-based localized contrast features based, at least in part, on a Michelson contrast value.

18. A non-transitory computer-readable storage medium carrying one or more sequences of one or more instructions which, when executed by one or more processors, cause an apparatus to at least perform the following steps:

determining image data from a plurality of bounding boxes each centering at an identical point of an image, wherein the bounding boxes are of different sizes each smaller than a size of the image;

initiating an extraction of one or more fogging attributes from the image data;

initiating a mapping of the one or more fogging attributes to one or more features of at least one classifier, wherein the at least one classifier is trained based, at least in part, on a co-registration of the one or more features to one or more records of previously made image judgements; and

initiating a classification of the image as either in a fogging-afflicted state or a non-fogging-afflicted state based, at least in part, on the at least one classifier.

19. A non-transitory computer-readable storage medium of claim 18,

wherein the bounding boxes are located within a center region of the image.

20. A non-transitory computer-readable storage medium of claim 19, wherein the identical point of the image is the center point of the image, the bounding boxes include, at least in part, one or more small bounding boxes, one or more medium bounding boxes, and one or more large bounding boxes specified for a plurality of color channels of the image.