Content-aware weighted image manipulations

Patent metadata

Field	Value
Publication number	US-9336567-B2
Application number	US-201414184620-A
Country	US
Kind code	B2
Filing date	Feb 19, 2014
Priority date	Dec 16, 2013
Publication date	May 10, 2016
Grant date	May 10, 2016

Abstract

A method in a computing device for performing intelligent weighted image manipulations is described. The method includes determining whether edge image features are distributed evenly across an image. When the edge image features in the image are not distributed evenly across the image, the method further includes cropping the image at the bounds of an overlay region of a desired size that is set at a position within the image to include a largest number of the edge image features. According to an embodiment, when the edge image features in the image are distributed evenly across the image, the method further includes cropping the image at the bounds of the overlay region of a desired size that is set at the center of the image.

First claim

What is claimed is:

1. A method in a computing device for performing intelligent weighted image manipulations, the method comprising:
determining whether edge image features are distributed evenly across an image, wherein the edge image features are not distributed evenly across the image when one of two non-overlapping regions created by placing two copies of an overlay region at opposite ends of the image has a first number of edge image features that exceeds a second number of edge image features of the other non-overlapping region by at least a threshold value; and
when the edge image features in the image are not distributed evenly across the image, cropping the image at the bounds of the overlay region of a desired size, wherein the overlay region is set at a position within the image to include a largest number of the edge image features, wherein the dimensions of the overlay region correspond to an aspect ratio of a display device that the cropped image is to be sent to, and wherein the dimensions are a largest proportional size, according to the aspect ratio, not exceeding the size of the image.

2. The method of claim 1 , further comprising:
when the edge image features in the image are distributed evenly across the image, cropping the image at the bounds of the overlay region of the desired size, wherein the overlay region is set at the center of the image.

3. The method of claim 1 , wherein the edge image features are locations of the image that include edges of elements represented within the image.

4. The method of claim 3 , further comprising:
modifying the cropped image to include a label located at one of a plurality of label overlay regions that bounds a least number of edge image features of the cropped image.

5. The method of claim 3 , wherein the image is one of a plurality of images of a video asset, wherein the method further comprises:
selecting a representative cropped image, from a plurality of cropped images corresponding to the plurality of images, which has a highest number of edge image features of all the plurality of cropped images.

6. The method of claim 1 , further comprising:
prior to the determining, orienting the image according to an orientation of a plurality of orientations that preserves a largest amount of the image within a boundary of the overlay region.

7. The method of claim 1 , wherein said determining whether edge image features are distributed evenly across an image comprises:
generating an edge detected image from the image, wherein the edge detected image includes a plurality of edge pixels indicating the edge image features at the respective pixel locations of the image; and
determining whether a first number of edge pixels within a first end region of the edge detected image exceeds a second number of edge pixels within a second end region of the edge detected image by at least the threshold value, wherein the second end region is a same size as the first end region and is located at an opposite side of the edge detected image from the first end region.

8. A media server to perform intelligent weighted image manipulations, comprising:
a memory storing instructions;
a media store to store images; and
a processor coupled with the media store and with the memory to execute the instructions to implement an image processing module to perform intelligent weighted image manipulations, the image processing module including, an intelligent aspect ratio cropping module to, for each of the images,
determine whether edge image features are distributed evenly across the image by determining whether a first number of edge pixels within a first end region of the edge detected image corresponding to the image exceeds a second number of edge pixels within a second end region of the edge detected image by at least a threshold value, wherein the second end region is a same size as the first end region and is located at an opposite side of the edge detected image from the first end region, and
when the edge image features in the image are not distributed evenly across the image, crop the image at the bounds of an overlay region of a desired size, wherein the overlay region is set at a position within the image to include a largest number of the edge image features; and

an edge detection module to generate edge detected images from the images, wherein each edge detected image includes a plurality of edge pixels indicating edge image features at the respective pixel locations of the respective image.

9. The media server of claim 8 , wherein the intelligent aspect ratio cropping module is further, for each of the images, to:
when the edge image features in the image are distributed evenly across the image, crop the image at the bounds of the overlay region of the desired size, wherein the overlay region is set at the center of the image.

10. The media server of claim 8 , wherein the image processing module further comprises a intelligent overlay module to, for one of the cropped images, modify the one cropped image to include a label located at one of a plurality of label overlay regions that bounds a least number of edge image features of the one cropped image.

11. The media server of claim 8 , wherein the image processing module further comprises a intelligent video frame selection module to:
select a representative cropped image, from the plurality of cropped images corresponding to a plurality of the images, which has a highest number of edge image features of all the plurality of cropped images.

12. A non-transitory computer-readable storage medium having instructions stored therein for performing intelligent weighted image manipulations, wherein the instructions, when executed by a processor of a media server, cause the processor to perform operations comprising:
determining whether edge image features are distributed evenly across an image, wherein the determining comprises:
generating an edge detected image from the image, wherein the edge detected image includes a plurality of edge pixels indicating the edge image features at the respective pixel locations of the image, and
determining whether a first number of edge pixels within a first end region of the edge detected image exceeds a second number of edge pixels within a second end region of the edge detected image by at least a threshold value, wherein the second end region is a same size as the first end region and is located at an opposite side of the edge detected image from the first end region; and

when the edge image features in the image are not distributed evenly across the image, cropping the image at the bounds of an overlay region of a desired size, wherein the overlay region is set at a position within the image to include a largest number of the edge image features.

13. The non-transitory computer-readable storage medium of claim 12 , wherein the operations further comprise:
when the edge image features in the image are distributed evenly across the image, cropping the image at the bounds of the overlay region of the desired size, wherein the overlay region is set at the center of the image.

14. The non-transitory computer-readable storage medium of claim 12 , wherein the operations further comprise:
modifying the cropped image to include a label located at one of a plurality of label overlay regions that bounds a least number of edge image features of the cropped image.

15. The non-transitory computer-readable storage medium of claim 12 , wherein the image is one of a plurality of images of a video asset, and wherein the operations further comprise:
selecting a representative cropped ima

Assignees

• Ericsson Telefon Ab L M

Inventors

• Lapointe Jean

Classifications

• G06T3/403 Primary

  Edge-driven scaling; Edge-based scaling · CPC title

• G06T3/40

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

• G06T11/60

  Creating or editing images; Combining images with text · CPC title

• G06T7/608

  Physics · mapped topic

• G06T7/68

  of symmetry · CPC title