Facilitating preservation of regions of interest in automatic image cropping
Embodiments of the present invention are directed to facilitating region of interest preservation. In accordance with some embodiments of the present invention, a region of interest preservation score using adaptive margins is determined. The region of interest preservation score indicates an extent to which at least one region of interest is preserved in a candidate image crop associated with an image. A region of interest positioning score is determined that indicates an extent to which a position of the at least one region of interest is preserved in the candidate image crop associated with the image. The region of interest preservation score and/or the preserving score are used to select a set of one or more candidate image crops as image crop suggestions.
1. A system comprising:
means for determining a region of interest preservation score using adaptive margins, the region of interest preservation score indicating an extent to which at least one region of interest is preserved within a valid region of a candidate image crop associated with an image, wherein determining the region of interest preservation score comprises determining whether the at least one region of interest falls within the valid region positioned within the candidate image crop, the valid region defined by adaptive margins corresponding with each boundary of the candidate image crop, the adaptive margins comprising an adaptive margin for each boundary of the candidate image crop that is selected from a smaller of a fixed-proportion margin and a distance between an aggregate region and an image boundary;
means for determining a region of interest positioning score that indicates an extent to which a position of the at least one region of interest in the image is preserved in the candidate image crop associated with the image; and
means for selecting the candidate image crop from among a set of candidate image crops based on at least one of the region of interest preservation score and the region of interest positioning score.
2. The system of claim 1 , further comprising determining for each boundary of the aggregate region, a distance between the aggregate region and the corresponding image boundary, the aggregate region including at least one region of interest in the image.
3. The system of claim 1 , wherein determining the region of interest positioning score comprises determining an extent to which an aggregate region maintains a relative position of the image in the candidate image crop as in the image.
4. The system of claim 1 further comprising determining a candidate score based on at least one of the region of interest preservation score and the region of interest positioning score.
5. The system of claim 4 further comprising selecting the candidate image crop from among the set of candidate image crops based on the candidate image score.
6. The system of claim 5 further comprising providing the selected candidate image crop as an image crop suggestion.
7. The system of claim 5 further comprising automatically cropping the image based on the selected candidate image crop.
8. A method for preserving regions of interest, the method comprising:
determining a region of interest preservation score using adaptive margins, the region of interest preservation score indicating an extent to which at least one region of interest is preserved in a candidate image crop associated with an image, wherein determining the region of interest preservation score comprises selecting an adaptive margin for each boundary of the candidate image crop that is smaller of a fixed-scale margin and a distance between an aggregate region and an image boundary; and
selecting the candidate image crop from among a set of candidate image crops based on at least the region of interest preservation score.
9. The method of claim 8 further comprising determining a candidate score based on at least the region of interest preservation score.
10. The method of claim 9 wherein the candidate score is used to rank the candidate image crop among other candidate image crops.
11. The method of claim 9 , wherein the candidate score is used to select the candidate image crop from among the set of candidate image crops.
12. The method of claim 11 , wherein the selected candidate image crop is provided as an image crop suggestion.
13. The method of claim 11 , wherein the selected candidate image crop is used to automatically crop the image.
14. One or more computer-readable media having a plurality of executable instructions embodied thereon, which, when executed by one or more processors, cause the one or more processors to perform a method comprising:
determining boundary distances between each boundary of an aggregate region and each corresponding boundary of an image, the aggregate region including at least one region of interest in the image;
for each boundary of a candidate image crop, selecting an adaptive margin inward from the corresponding boundary that defines a valid region, the selected adaptive margin for each boundary being a minimum of the boundary distance associated with the boundary and a fixed-scale margin;
determining a region of interest preservation score using the adaptive margin, the region of interest preservation score indicating an extent to which the at least one region of interest is preserved within the valid region, wherein determining the region of interest preservation score comprises determining whether the at least one region of interest falls within the valid region defined by selected margins for each boundary of the candidate image crop; and
based on the region of interest preservation score, automatically providing the candidate image crop as an image crop suggestion that, when selected, is used to automatically crop the image.
15. The one or more computer-readable media of claim 14 , the method further comprising determining the aggregate region based on the at least one region of interest.
16. The one or more computer-readable media of claim 14 , the method further comprising generating a candidate score for the candidate image crop utilizing the extent to which the at least one region of interest is maintained in the candidate image crop.
17. The one or more computer-readable media of claim 16 , the method further comprising determining the candidate score based on at least the region of interest preservation score.
18. The one or more computer-readable media of claim 17 , the method further comprising using the candidate score to rank the candidate image crop among other candidate image crops, and selecting a set of candidate image crops for providing as image crop suggestions.
19. The one or more computer-readable media of claim 17 , wherein the candidate score is used to select the candidate image crop from among the set of candidate image crops.
20. The one or more computer-readable media of claim 19 , wherein the selected candidate image crop is provided as an image crop suggestion.