Document-occluding artifact removal

1 . A non-transitory computer-readable data storage medium storing program code executable by a processor to perform processing comprising:

determining depth information for each pixel of a plurality of pixels of a digitally captured image of a document occluded by an object;

identifying boundaries of the document within the digitally captured image;

determining a baseline depth of the document from the depth information for the pixels inside the document boundaries;

identifying the pixels for which the depth information is indicative of a greater depth than the baseline depth of the document as belonging to an artifact corresponding to the object occluding the document; and

removing the artifact from the digitally captured image.

2 . The non-transitory computer-readable data storage medium of claim 1 , wherein the processing further comprises:

performing optical character recognition (OCR) on the document within the digitally captured image from which the artifact has been removed.

3 . The non-transitory computer-readable data storage medium of claim 1 , wherein identifying the boundaries of the document within the digitally captured image comprises:

segmenting the image into a plurality of regions without using the depth information;

identifying the segmented region corresponding to the document, boundaries of the segmented region corresponding to the document as the boundaries of the document within the digitally captured image.

4 . The non-transitory computer-readable data storage medium of claim 3 , wherein the segmented region corresponding to the document is polygonal in shape, each boundary of the segmented region is linear, and pairs of adjacent boundaries of the segmented region each meet at a corner.

5 . The non-transitory computer-readable data storage medium of claim 1 , wherein determining the baseline depth of the document from the depth information for the pixels inside the document boundaries comprises:

determining the baseline depth as a most frequent depth indicated by the depth information for the pixels inside the document boundaries, within a threshold.

6 . The non-transitory computer-readable data storage medium of claim 1 , wherein identifying the pixels for which the depth information is indicative of the greater depth than the baseline depth of the document as belonging to the artifact comprises:

identifying the pixels inside the document boundaries and for which the depth information is indicative of the greater depth as belonging to the artifact.

7 . The non-transitory computer-readable data storage medium of claim 6 , wherein identifying the pixels for which the depth information is indicative of the greater depth than the baseline depth of the document as belonging to the artifact further comprises:

identifying the pixels outside the document boundaries and for which the depth information is indicative of the greater depth as belonging to the artifact.

8 . The non-transitory computer-readable data storage medium of claim 1 , wherein the processing further comprises:

adjusting the document boundaries to include groups of the pixels outside the boundaries, which are adjacent to the document within the digitally captured image, and for which the depth information is indicative of the baseline depth of the document.

9 . The non-transitory computer-readable data storage medium of claim 1 , wherein the processing further comprises:

adjusting the document boundaries to exclude groups of the pixels inside the boundaries and for which the depth information is indicative of a lower depth than the baseline depth of the document.

10 . The non-transitory computer-readable data storage medium of claim 1 , wherein removing the artifact from the digitally captured image comprises:

setting each pixel belonging to the artifact to a minimum pixel value.

11 . The non-transitory computer-readable data storage medium of claim 10 , wherein removing the artifact from the digitally captured image comprises:

inpainting groups of the pixels set to the minimum pixel value.

12 . A computing device comprising:

image-capturing hardware to digitally capture an image of a document occluded by an object, the image having a plurality of pixels;

a processor; and

a memory storing program code executable by the processor to:

determine depth information for each pixel of the image;

identify boundaries of the document within the image;

determine a baseline depth of the document from the depth information for the pixels inside the document boundaries; and

remove the pixels from the image for which the depth information is indicative of a greater depth than the baseline depth of the document.

13 . The computing device of claim 12 , wherein the program code is executable by the processor to further:

perform optical character recognition (OCR) on the document within the image from which the pixels have been removed.