Image resolution enhancement based on data from related images

1. A method to enhance image resolution of a cropped image using a set of related images, the method comprising:

dividing, by one or more processors, the cropped image and each of the related images into blocks that each include a subset of pixels of a respective image, one or more of the blocks in the related images having one or more features that match one or more features of respective counterpart blocks in the cropped image;

determining, by the one or more processors, blocks in the related images that have a relatively higher resolution than the respective counterpart blocks in the cropped image; and

replacing, by the one or more processors, one or more of the counterpart blocks in the cropped image with respective determined blocks from the related images to enhance the image resolution of at least a portion of the cropped image.

2. A method as recited in claim 1 , further comprising selecting the set of related images from a plurality of images by filtering the plurality of images based on metadata associated with the cropped image.

3. A method as recited in claim 2 , wherein the metadata associated with the cropped image includes one or more of a capture date or a global position system (GPS) coordinates associated with the cropped image, the capture date representing a date that the cropped image was captured by an image capturing device, and the GPS setting representing a physical location where the cropped image was captured.

4. A method as recited in claim 1 , further comprising extracting features from the cropped image or the related images by at least transforming the cropped image or the related images into a collection of local feature vectors that are invariant to one or more of scaling, rotation, or translation of the cropped image or the related images.

5. A method as recited in claim 1 , further comprising partitioning the cropped image and each of the related images into the blocks based on locations of features in each said image and pixel intensity differences proximal to the locations.

6. A method as recited in claim 1 , wherein said replacing includes projecting one or more pixels from the respective determined blocks in the related images onto the one or more of the counterpart blocks in the cropped image.

7. A method as recited in claim 1 , further comprising:

detecting features in the cropped image;

extracting features from the related images; and

mapping the features detected in the cropped image to matching features extracted from the related images.

8. A method as recited in claim 1 , wherein the determining comprises:

mapping a counterpart block in the cropped image to a block in a single related image; and

selecting the block from the single related image for replacement of said counterpart block in the cropped image based on the block having a size that is at least one and a half times a size of said counterpart block.

9. A method as recited in claim 1 , wherein the determining comprises:

mapping a counterpart block in the cropped image to multiple related images; and

selecting a particular block from one of the multiple related images for replacement of the counterpart block in the cropped image based on the particular block having:

at least a maximum number of features matching the one or more features in the counterpart block from the cropped image in comparison to additional blocks from the multiple related images that include one or more features that match the one or more features in the counterpart block from the cropped image; and

a relatively largest size in comparison to respective sizes of the additional blocks from the multiple related images.

10. A method as recited in claim 1 , wherein the replacing comprises:

creating a projection matrix using locations corresponding to features in a counterpart block in the cropped image; and

projecting the subset of pixels from a determined block in a related image into the counterpart block in the cropped image to correspond to an original alignment in the cropped image based on the projection matrix.

11. A method as recited in claim 1 , further comprising:

cropping an image to create the cropped image;

enlarging the cropped image effective to cause pixilation and create an enlarged pixilated version of the cropped image; and

receiving a request to enhance image resolution of the enlarged pixilated version of the cropped image using the set of related images.

12. A system, comprising:

one or more processors; and

a memory having instructions that are executable by the one or more processors to implement an image service that is configured to enhance image resolution of an input image using a set of related images, the image service configured to:

divide the input image and the related images into blocks that each contain a subset of pixels of a respective image;

determine the blocks in the related images that have a relatively higher resolution than respective counterpart blocks in the input image; and

project pixels from the determined blocks in the related images onto respective counterpart blocks in the input image to generate an output image that has enhanced image resolution in comparison to the input image.

13. A system as recited in claim 12 , wherein the image service further comprises a filter module configured to filter a plurality of images based on metadata associated with the input image to identify the set of related images from among the plurality of images.

14. A system as recited in claim 12 , wherein the image service further comprises a feature matching module that is configured to map features in the input image to matching features in one or more of the related images.

15. A system as recited in claim 12 , wherein the image service is configured to determine the blocks in the related images that have a relatively higher resolution than respective counterpart blocks in the input image based on:

at least a maximum number of features that match one or more features in the respective counterpart block from the input image in comparison to additional blocks from the related images that include one or more features that match the one or more features in the counterpart block from the input image; and

a relatively largest size in comparison to respective sizes of the additional blocks from the related images.

16. Computer-readable storage memory comprising stored instructions that are executable by a computing device to implement an image service configured to perform operations comprising:

dividing a cropped image and each related image from a set of related images into blocks that each include a subset of pixels of a respective image, one or more of the blocks in the related images having one or more features that match one or more features of respective counterpart blocks in the cropped image;

determining blocks in the related images that have a relatively higher resolution than the respective counterpart blocks in the cropped image; and

replacing one or more of the counterpart blocks in the cropped image with respective determined blocks from the related images to enhance image resolution of at least a portion of the cropped image.

17. Computer-readable storage memory as recited in claim 16 , wherein the operations further comprise selecting the set of related images from a plurality of images by filtering the plurality of images based on metadata associated with the cropped image.

18. Computer-readable storage memory as recited in claim 16 , wherein the metadata associated with the cropped image includes one or more of a capture date or a global position system (GPS) coordinates associated with the cropped image, the capture date representing a date that the cropped image was captured by an image capturing device, and the GPS setting representing a location where the cropped image was captured.

19. Computer-readable storage memory as recited in claim 16 , wherein the operations further comprise extracting features from the cropped image or the related images by at least transforming the cropped image or the related images into a collection of local feature vectors that are invariant to one or more of scaling, rotation, or translation of the cropped image or the related images.

20. Computer-readable storage memory as recited in claim 16 , wherein said replacing includes projecting one or more pixels from the respective determined blocks in the related images onto the one or more of the counterpart blocks in the cropped image.