System and method for capturing relevant information from a printed document
View Patent ↗A city directory, having a listing of names and associated information of residents in a city (or similar location), is digitized. Zones of text having information not useful to users of the digitized directory are removed, and lines of information corresponding to residents are reconstructed, to make the digitized directory more easily accessed and reviewed.
1. A computerized method for capturing relevant information from a printed document, the document having at least some information arranged in lines, some of the arranged information being relevant and other of the arranged information being irrelevant for purposes of a user, the method comprising:
receiving the arranged information on the document after it has been optically scanned;
using character recognition to convert the scanned information into computer readable text;
defining zones of the document, with the computer readable text assigned to one of the zones;
applying rules to the zones of the document, in order to identify information in the zones that is not relevant;
excluding any zone of the document having information that is not relevant;
applying rules to the information remaining after the step of excluding, the rules identifying a line in a remaining zone that is associated with another line; and
combining the indentified line and its associated line into a reconstructed line.
2. The method of claim 1 , wherein the scanned information is in the form of images, and wherein an optical character recognition (OCR) system is used to convert the scanned information into computer readable text.
3. The method of claim 2 , wherein the zones of the document are text zones, and wherein the OCR system creates the text zones from, at least in part, analysis of the computer readable text converted by the OCR system from the scanned information.
4. The method of claim 3 , wherein the step of excluding includes removing a text zone from the document based on one or more of zone size characteristics, zone format characteristics, lower case characteristics, index characteristics, paragraph characteristics, and address/number first characteristics.
5. The method of claim 4 , wherein:
zone size characteristics comprise having have three or fewer lines in a zone or a the X or horizontal dimension of a zone being less than ten characters.
6. The method of claim 4 , wherein:
zone format characteristics comprise more than 25% of the lines in a zone varying more than 50% from the median line height for the document.
7. The method of claim 4 , wherein:
lower case characteristics comprise a majority of lines in a zone beginning with a lower case word.
8. The method of claim 4 , wherein:
index characteristics comprise a pattern of dots or periods in a zone leading to a number.
9. The method of claim 4 , wherein:
paragraph characteristics comprise indented lines in a zone, where the first letter of each indented line is capped, and where each such indented line is followed by one or more lines that have a full left alignment.
10. The method of claim 4 , wherein:
Address/number first characteristics comprise more than 25% of the lines in a zone beginning with a number.
11. The method of claim 1 , further comprising:
determining if any line of information has a ditto designation; and
adding a ditto word from a preceding line to any line having a ditto designation.
12. The method of claim 1 , wherein a plurality of indentified lines are combined with associated lines to provide a plurality of reconstructed lines.
13. The method of claim 1 , wherein the printed document is a historical record.
14. The method of claim 13 , wherein the historical record is a city directory.
15. The method of claim 1 , wherein the relevant arranged information comprises one or more of names, addresses, and occupations.
16. The method of claim 1 , wherein the irrelevant arranged information comprises one or more of advertising, lists of lower case words and abbreviations, indexes, information text, address first listings, and number first listings.
17. A document processing system for capturing relevant information from a printed document, the document having at least some information arranged in lines, some of the arranged information being relevant and other of the arranged information being irrelevant for purposes of a user, the system comprising one or more processors, the one or more processors configured to:
receive the arranged information on the document after it has been scanned;
use character recognition to convert the scanned information into computer readable text;
define zones of the document, with the computer readable text assigned to one of the zones;
apply rules to the zones of the document, in order to identify information in the zones that is not relevant;
exclude any zone of the document having information that is not relevant;
apply rules to the information remaining after the step of excluding, the rules identifying a line in a remaining zone that is associated with another line; and
combine the indentified line and its associated line into a re-constructed line.
18. The system of claim 17 , wherein the scanned information is in the form of images, and wherein one of the processors is part of an optical character recognition (OCR) system that is used to convert the scanned information into computer readable text.
19. The system of claim 18 , wherein the zones of the document are text zones, and wherein the OCR system creates the text zones from, at least in part, analysis of the computer readable characters converted by the OCR system from the scanned information.
20. The system of claim 17 , wherein the one or more processors are further configured to exclude zones by removing a text zone from the document based on one or more of zone size characteristics, zone format characteristics, lower case characteristics, index characteristics, paragraph characteristics, and address/number first characteristics.
21. The system of claim 17 , wherein the one or more processors are further configured to:
determine if any line of information has a ditto designation; and
add a ditto word from a preceding line to any line having a ditto designation.
22. The system of claim 17 , wherein the one or more processors are further configured to combine a plurality of indentified lines with associated lines to provide a plurality of reconstructed lines.
23. The system of claim 17 , wherein the printed document is a historical record.
24. The system of claim 23 , wherein the historical record is a city directory.
25. The system of claim 17 , wherein the relevant arranged information comprises one or more of names, addresses, and occupations.
26. The system of claim 17 , wherein the irrelevant arranged information comprises one or more of advertising, lists of lower case words and abbreviations, indexes, information text, address first listings, and number first listings.