Page layout determination of an image undergoing optical character recognition
A method and system is provided for identifying a page layout of an image that includes textual regions. The textual regions are to undergo optical character recognition (OCR). The system includes an input component that receives an input image that includes words around which bounding boxes have been formed and a text identifying component that groups the words into a plurality of text regions. A reading line component groups words within each of the text regions into reading lines. A text region sorting component that sorts the text regions in accordance with their reading order.
1. A method performed on a computing device including a processor and memory, the method comprising:
identifying, by the computing device, whitespace regions in a digital image;
grouping, by the computing device, words in the digital image into text regions in the digital image that do not include any of the identified whitespace regions, where the text regions each comprise at least a portion of the words arranged in text lines that are in a vertical order;
determining, by the computing device based on the vertical orders of the text lines in each text region, a reading order of the text regions; and
sorting, by the computing device, the text regions in accordance with the reading order.
2. The method of claim 1 further comprising bonding each of the words in a bounding box.
3. The method of claim 1 where the whitespace regions include any images positioned among the words.
4. The method of claim 1 where none of the identified whitespace regions overlap each other.
5. The method of claim 1 further comprising identifying the text lines within each of the text regions.
6. The method of claim 1 where space between columns of the words is encompassed by the whitespace regions.
7. The method of claim 1 where the reading order comprises a natural flow of text encompassed by the sorted text regions.
8. A system comprising:
a processor;
memory coupled to the processor;
a white space component implemented at least in part by the processor and memory, and configured to identify whitespace regions in a digital image;
a reading line component implemented at least in part by the processor and memory, and configured to group words in the digital image into text regions in the digital image that do not include any of the identified whitespace regions, where the text regions each comprise at least a portion of the words arranged in text lines that are in a vertical order;
a reading line component further configured to determine, based on the vertical orders of the text lines in each text region, a reading order of the text regions; and
a text region sorting component implemented at least in part by the processor and memory, and configured to sort the text regions in accordance with the reading order.
9. The system of claim 8 further comprising a text region formation component implemented at least in part by the processor and memory, and configured to bound each of the words in a bounding box.
10. The system of claim 9 where the whitespace regions include any images positioned among the words while not overlapping any of the bounding boxes.
11. The system of claim 8 where none of the identified whitespace regions overlap each other.
12. The system of claim 8 , the actions further comprising creating reading lines within each of the final text regions.
13. The system of claim 8 where space between columns of the words is encompassed by the whitespace regions.
14. The system of claim 8 where the reading order comprises a natural flow of text of the sorted text regions.
15. At least one computer storage medium comprising:
a hardware memory that comprises software that, based on execution by a computing device that includes a processor and memory, configures the computing device to perform actions comprising:
identifying, by the computing device, whitespace regions in a digital image;
grouping, by the computing device, words in the digital image into text regions in the digital image that do not include any of the identified whitespace regions, where the text regions each comprise at least a portion of the words arranged in text lines that are in a vertical order;
determining, by the computing device based on the vertical orders of the text lines in each text region, a reading order of the text regions; and
sorting, by the computing device, the text regions in accordance with the reading order.
16. The at least one computer storage medium of claim 15 the actions further comprising bounding each of the words in a bounding box.
17. The at least one computer storage medium of claim 15 where none of the identified whitespace regions overlap each other.
18. The at least one computer storage medium of claim 15 where the text regions comprise inter-word spacings but not inter-column spacings.
19. The at least one computer storage medium of claim 16 where the whitespace regions include any images positioned among the words while not overlapping any of the bounding boxes.
20. The at least one computer storage medium of claim 15 where space between columns of the words is encompassed by the whitespace regions, or where the reading order comprises a natural flow encompassed by text of the sorted text regions.