Extracting multiple documents from single image
System and method for document image detection, comprising: producing, using a neural network, a superpixel segmentation map of an input image; generating a superpixel binary mask by associating each superpixel of the superpixel segmentation map with a class of a predetermined set of classes; identifying one or more connected components in the superpixel binary mask; for each connected component of the superpixel binary mask, identifying a corresponding minimum bounding polygon; creating one or more image dividing lines based on the minimum bounding polygons; and defining boundaries of one or more objects of interest based on at least a subset of the image dividing lines.
1. A computer-implemented method for document image detection, comprising:
generating a superpixel binary mask associated with an input image, wherein each superpixel of the superpixel binary mask is derived from a probability characteristic reflecting a probability of the superpixel belonging to a certain object found in an input image;
identifying a connected component in the superpixel binary mask;
responsive to determining that a first number of pixels in a first line of the superpixel binary mask exceeds, by at least a predetermined threshold, a second number of pixels in a second line of the superpixel binary mask which is adjacent to the first line of the superpixel binary mask, utilizing the second line as a candidate boundary of a minimum bounding polygon associated with the connected component;
creating one or more image dividing lines based on the minimum bounding polygon; and
defining boundaries of one or more regions of interest based on at least a subset of the image dividing lines.
2. The method of claim 1 , wherein the superpixel binary mask is generated based on a chosen binarization threshold.
3. The method of claim 1 , wherein the first line is provided by one of: a row of the superpixel binary mask or a column of the superpixel binary mask.
4. The method of claim 1 , wherein identifying the minimum bounding polygon further comprises:
generating a plurality of candidate lines for the minimum bounding polygon;
computing a value of a quality metric for the one or more regions of interest.
5. The method of claim 4 , wherein generating the plurality of candidate lines for the minimum bounding polygon further comprises:
responsive to determining that a first number of pixels in a first line of the superpixel binary mask exceeds, by at least a predetermined threshold, a second number of pixels in a second line of the superpixel binary mask which is adjacent to the first line of the superpixel binary mask, utilizing the second line as a second candidate boundary of the bounding polygon, wherein the first line is provided by one of: a row of the superpixel binary mask or a column of the superpixel binary mask.
6. The method of claim 4 , wherein generating the plurality of candidate lines for the minimum bounding polygon further comprises:
utilizing, as a second candidate boundary of the bounding polygon, a line traversing a center of the superpixel binary mask.
7. The method of claim 4 , wherein computing a value of a quality metric for one or more regions of interest further comprises:
applying, to the one or more regions of interest, a trainable classifier.
8. A system, comprising:
a memory;
a processor, coupled to the memory, the processor configured to:
generate a superpixel binary mask associated with an input image, wherein each superpixel of the superpixel binary mask is derived from a probability characteristic reflecting a probability of the superpixel belonging to a certain object found in an input image;
identify a connected component in the superpixel binary mask;
responsive to determining that a first number of pixels in a first line of the superpixel binary mask exceeds, by at least a predetermined threshold, a second number of pixels in a second line of the superpixel binary mask which is adjacent to the first line of the superpixel binary mask, utilize the second line as a candidate boundary of a minimum bounding polygon associated with the connected component;
create one or more image dividing lines based on the minimum bounding polygon; and
define boundaries of one or more regions of interest based on at least a subset of the image dividing lines.
9. The system of claim 8 , wherein the superpixel binary mask is generated based on a chosen binarization threshold.
10. The system of claim 8 , wherein the first line is provided by one of: a row of the superpixel binary mask or a column of the superpixel binary mask.
11. The system of claim 8 , wherein identifying the minimum bounding polygon further comprises:
generating a plurality of candidate lines for the minimum bounding polygon;
computing a value of a quality metric for the one or more regions of interest.
12. The system of claim 11 , wherein generating the plurality of candidate lines for the minimum bounding polygon further comprises:
responsive to determining that a first number of pixels in a first line of the superpixel binary mask exceeds, by at least a predetermined threshold, a second number of pixels in a second line of the superpixel binary mask which is adjacent to the first line of the superpixel binary mask, utilizing the second line as a second candidate boundary of the bounding polygon, wherein the first line is provided by one of: a row of the superpixel binary mask or a column of the superpixel binary mask.
13. The system of claim 11 , wherein generating the plurality of candidate lines for the minimum bounding polygon further comprises:
utilizing, as a second candidate boundary of the bounding polygon, a line traversing a center of the superpixel binary mask.
14. The system of claim 11 , wherein computing a value of a quality metric for one or more regions of interest further comprises:
applying, to the one or more regions of interest, a trainable classifier.
15. A non-transitory computer-readable storage medium comprising executable instructions that, when executed by a computer system, cause the computer system to:
generate a superpixel binary mask associated with an input image, wherein each superpixel of the superpixel binary mask is derived from a probability characteristic reflecting a probability of the superpixel belonging to a certain object found in an input image;
identify a connected component in the superpixel binary mask;
responsive to determining that a first number of pixels in a first line of the superpixel binary mask exceeds, by at least a predetermined threshold, a second number of pixels in a second line of the superpixel binary mask which is adjacent to the first line of the superpixel binary mask, utilize the second line as a candidate boundary of a minimum bounding polygon associated with the connected component;
create one or more image dividing lines based on the minimum bounding polygon; and
define boundaries of one or more regions of interest based on at least a subset of the image dividing lines.
16. The non-transitory computer-readable storage medium of claim 15 , wherein the superpixel binary mask is generated based on a chosen binarization threshold.
17. The non-transitory computer-readable storage medium of claim 15 , wherein the first line is provided by one of: a row of the superpixel binary mask or a column of the superpixel binary mask.
18. The non-transitory computer-readable storage medium of claim 15 , wherein identifying the minimum bounding polygon further comprises:
generating a plurality of candidate lines for the minimum bounding polygon;
computing a value of a quality metric for the one or more regions of interest.
19. The non-transitory computer-readable storage medium of claim 18 , wherein generating the plurality of candidate lines for the minimum bounding polygon further comprises:
responsive to determining that a first number of pixels in a first line of the superpixel binary mask exceeds, by at least a predetermined threshold, a second number of pixels in a second line of the superpixel binary mask which is adjacent to the first line of the superpixel binary mask, utilizing the second line as a second candidate boundary of the bounding polygon, wherein the first line is provided by one of: a row of the superpixel binary mask or a column of the superpixel binary mask.
20. The non-transitory computer-readable storage medium of claim 18 , wherein generating the plurality of candidate lines for the minimum bounding polygon further comprises:
utilizing, as a second candidate boundary of the bounding polygon, a line traversing a center of the superpixel binary mask.