Interactive refocusing of electronic images

1. A method of refocusing an image, the method comprising:

in a processor, obtaining a set of refocused images generated from a single captured light field image comprising a plurality of pixels and representing a scene, wherein each refocused image comprises an image of the scene at a distinct focal depth;

in the processor, determining refocus depths for at least a portion of the captured image;

storing the refocus depths in a look-up table; and

in the same or a different processor, refocusing at least a portion of the captured image at a desired refocus depth determined from the look-up table;

wherein determining refocus depths for at least a portion of the captured image comprises, for each of at least a subset of the pixels, determining the refocus depth for the pixel by searching other pixels surrounding the pixel until a pixel with a known refocus depth is found.

2. The method of claim 1 , wherein the refocus depths span a range of depths including closest and furthest distances to refocus.

3. The method of claim 2 , wherein the single captured light field image comprises an image captured by a light field capture device, and wherein the range of depths corresponds to a depth of focus of the light field capture device.

4. The method of claim 3 , wherein the range of depths is extended beyond an effective depth of focus of the light field capture device by a factor representing a tolerable mis-focus.

5. The method of claim 2 , wherein the furthest distance is optical infinity.

6. The method of claim 2 , wherein the closest distance corresponds to a closest object in the scene, and wherein the furthest distance corresponds to a furthest object in the scene.

7. The method of claim 1 , wherein refocus depths for at least a subset of pixels in the plurality of pixels.

8. The method of claim 1 , wherein determining refocus depths comprises determining refocus depths for respective regions in the captured image.

9. The method of claim 1 , further comprising downsampling the look-up table.

10. The method of claim 9 , wherein the downsampling operation includes a voting procedure wherein pixel votes are weighted.

11. The method of claim 1 , further comprising displaying the captured image on a display device, and wherein the desired refocus depth corresponds to a first refocus depth associated with a pixel at a user-specified location in the displayed image.

12. The method of claim 11 , wherein refocusing at least a portion of the captured image comprises refocusing a region of the displayed image that includes the user-specified location.

13. The method of claim 12 , wherein the region is user-defined.

14. The method of claim 12 , wherein the region corresponds to an object in the displayed image.

15. The method of claim 12 , wherein a tag is associated with the region.

16. The method of claim 1 , further comprising displaying the captured image on a display device, and wherein refocusing at least a portion of the captured image comprises generating an animation comprising a series of refocused images in the set having depths between a depth of the displayed image and the desired refocus depth.

17. The method of claim 1 , wherein refocusing at least a portion of the captured image comprises blending at least a portion of two refocused images in the set focused at depths near the desired refocus depth.

18. The method of claim 1 , further comprising storing the set of refocused images and the look-up table in a single file.

19. The method of claim 18 , wherein the single file comprises one of a zip archive and a swf (Flash) file.

20. The method of claim 1 , wherein obtaining a set of refocused images comprises:

receiving a captured light field image comprising light field data representing a scene;

determining a first depth corresponding to a nearest object in the scene;

determining a second depth corresponding to a farthest object in the scene;

digitally refocusing the light field image to generate a plurality of refocused images, each having a focus depth lying within the range from the first depth to the second depth;

storing the generated plurality of refocused images in a storage device; and

for each of a plurality of two-dimensional positions in the light field image:

determining which of the plurality of refocused images is in focus at the two-dimensional position; and

storing, in the look-up table, an index entry associating the two-dimensional position with the identified refocused image;

and wherein refocusing at least a portion of the captured image at a desired refocus depth comprises:

displaying one of the refocused images, the displayed image having a first focus depth;

receiving user input to select a two-dimensional position within the displayed image; and

in response to the received input:

retrieving, from the look-up table, an index entry identifying a stored refocused image corresponding to the selected two-dimensional position, the stored refocused image having a second focus depth different from the first focus depth;

retrieving the identified refocused image from the storage device; and

displaying the retrieved refocused image having the second focus depth.

21. The method of claim 20 , wherein digitally refocusing the light field image to generate a plurality of refocused images comprises:

scoring the set of images using a metric based on image content; and

removing at least one image from the set, based on the metric.

22. A system for refocusing an image, the system comprising:

a processor, configured to perform the operations of:

obtaining a set of refocused images generated from a single captured light field image comprising a plurality of pixels and representing a scene, wherein each refocused image comprises an image of the scene at a distinct focal depth; and

determining refocus depths for at least a portion of the captured image; and

a storage device, communicatively coupled to the processor, configured to store the refocus depths in a look-up table;

wherein the processor is further configured to refocus at least a portion of the captured image at a desired refocus depth determined from the look-up table;

wherein determining refocus depths for at least a portion of the captured image comprises, for each of at least a subset of the pixels, determining the refocus depth for the pixel by searching other pixels surrounding the pixel until a pixel with a known refocus depth is found.

23. The system of claim 22 , wherein the refocus depths span a range of depths including closest and furthest distances to refocus.

24. The system of claim 23 , wherein the closest distance corresponds to a closest object in the scene, and wherein the furthest distance corresponds to a furthest object in the scene.

25. The system of claim 22 , wherein the captured image comprises a plurality of pixels, and wherein determining refocus depths comprises determining refocus depths for at least a subset of pixels in the plurality of pixels.

26. The system of claim 22 , wherein determining refocus depths comprises determining refocus depths for respective regions in the captured image.

27. The system of claim 22 , wherein the storage device stores a downsampled look-up table.

28. The system of claim 22 , further comprising:

a display device, communicatively coupled to the processor, configured to display the captured image; and

an input device, communicatively coupled to the processor, configured to receive user input;

and wherein the desired refocus depth corresponds to a first refocus depth associated with a pixel at a location in the displayed image, wherein the location is specified by a user via the input device.

29. The system of claim 22 , further comprising:

a sensor, communicatively coupled to the processor, configured to receive a light field image comprising light field data representing a scene;

a display device, communicatively coupled to the processor, configured to display the captured image; and

an input device, communicatively coupled to the processor, configured to receive user input;

and wherein:

the processor is configured to obtain the set of refocused images by:

receiving a captured light field image comprising light field data representing a scene;

determining a first depth corresponding to a nearest object in the scene;

determining a second depth corresponding to a farthest object in the scene; and

digitally refocusing the light field image to generate a plurality of refocused images, each having a focus depth lying within the range from the first depth to the second depth; and

the storage device is configured to store the generated plurality of refocused images;

and wherein:

for each of a plurality of two-dimensional positions in the light field image:

the processor is further configured to determine which of the plurality of refocused images is in focus at the two-dimensional position; and

the storage device is configured to store, in the look-up table, an index entry associating the two-dimensional position with the identified refocused image;

the display is configured to display one of the refocused images, the displayed image having a first focus depth; and

the input device is configured to receive user input to select a two-dimensional position within the displayed image;

and wherein the processor is configured to refocus at least a portion of the captured image at a desired refocus depth by, in response to the received input:

retrieving, from the look-up table, an index entry identifying a stored refocused image corresponding to the selected two-dimensional position, the stored refocused image having a second focus depth different from the first focus depth;

retrieving the identified refocused image from the storage device; and

causing the display device to display the retrieved refocused image having the second focus depth.

30. The system of claim 29 , wherein the processor is configured to digitally refocus the light field image to generate a plurality of refocused images by:

scoring the set of images using a metric based on image content; and

removing at least one image from the set, based on the metric.