Ruled line extracting program, ruled line extracting apparatus and ruled line extracting method

• US 20080056576A1
• Filed: 11/30/2006
• Published: 03/06/2008
• Est. Priority Date: 08/31/2006
• Status: Active Grant

• Current Assignee: Fujitsu Limited
• Original Assignee: Fujitsu Limited
• Inventors: Takebe, Hiroaki, Fujimoto, Katsuhito
• Granted Patent: US 7,769,234 B2
• Time in Patent Office: Days
• US Class Current: 382/180
• CPC Class Codes: G06V 30/10 Character recognition
• CPC Class Codes: G06V 30/155 Removing patterns interfering with the pattern to be recognised, such as ruled lines or underlines
• CPC Class Codes: G06V 30/412 Layout analysis of documents structured with printed lines or input boxes, e.g. business forms or tables

Abstract

A ruled line extracting apparatus, a ruled line extracting program and a ruled line extracting method re-extract a ruled line by changing the predetermined requirements to be met by ruled line s when a ruled line candidate extracted according to the requirements shows a low reliability. A ruled line extracting program that causes a computer to extract a ruled line in an image of a document comprises an extraction step that extracts a ruled line candidate from the image of a document according to the first requirement predefined to be met by the figures of the elements of the ruled lines, a judgment step that judges if the ruled line candidate is stable or unstable according to the structural stability of the ruled line candidate extracted in the extraction step, a requirement determination step that determines the second requirement to be met by the figures of the elements of the ruled line different from the first requirement according to the ruled line candidate judged as stable in the judgment step and the first requirement and a re-extraction step that re-extracts a ruled line candidate according to the second requirement determined in the requirement determination step.

Claims

1. A ruled line extracting program that causes a computer to extract a ruled line in an image of a document, comprising:

an extraction step that extracts a ruled line candidate from the image of a document according to a first requirement predefined to be met by figures of elements of the ruled lines;

a judgment step that judges if the ruled line candidate is stable or unstable according to a structural stability of the ruled line candidate extracted in the extraction step;

a requirement determination step that determines a second requirement to be met by the figures of the elements of the ruled line different from the first requirement according to the ruled line candidate judged as stable in the judgment step and the first requirement; and

a re-extraction step that re-extracts a ruled line candidate according to the second requirement determined in the requirement determination step.

2. The program according to claim 1, wherein

the first requirement includes a parameter range showing the range of an element parameter that indicates properties of a graphic to be selected as a ruled line element of the ruled line and/or the range of a consolidation parameter that indicates properties of the graphic of the ruled line element to be consolidated as ruled line candidate.

3. The program according to claim 2, wherein

the extraction step is adapted to select a graphic that makes the element parameter to be found within the corresponding parameter range from the document image and consolidate a plurality of ruled line elements that make the consolidation parameter to be found within the corresponding parameter range to generate the ruled line candidate.

4. The program according to claim 3, wherein

the ruled line is a broken ruled line;

the parameter range of the element parameter is defined by a threshold value for sizes of circumscribed rectangles of black pixel link components; and

the extraction step being adapted to extract the black pixel link components from a document image expressed by a binary system and select the black pixel link components that make the element parameter to be found within the corresponding parameter range as ruled line elements.

5. The program according to claim 3, wherein

the ruled line is a broken ruled line;

the parameter range of the consolidation parameter is defined by a threshold value for a gap separating two ruled line elements, a threshold value for a drift in the direction perpendicular to the direction of consolidating two ruled line elements and/or a threshold value for the number of consolidated ruled line elements; and

the extraction step being adapted to select a plurality of ruled line elements that make the consolidation parameter to be found within the parameter range as ruled line candidate.

6. The program according to claim 3, wherein

the ruled line is a solid ruled line;

and

the extraction step is adapted to divide a document image expressed by a binary system into a plurality of regions, select regions that make the element parameter to be found within the parameter range out of the regions as ruled line elements and consolidate the ruled line elements to generate a ruled line candidate.

7. The program according to claim 2, wherein

the judgment step defines a third requirement by changing the parameter range of the first requirement according to the parameters of the ruled line elements constituting the ruled line candidate, extracts an interim ruled line candidate from the document image according to the third requirement as in the extraction step and judges the ruled line candidate extracted in the extraction step as unstable when it differs from the interim ruled line candidate.

8. The program according to claim 2, wherein

the judgment step defines the third requirement by changing the parameter range of the first requirement according to the parameters of the ruled line elements constituting the ruled line candidate, extracts an interim ruled line candidate from the document image according to the third requirement as in the extraction step and judges the ruled line candidate extracted in the extraction step as unstable when any of the ruled line elements constituting the interim ruled line candidate is not larger than a predetermined threshold value.

9. The program according to claim 2, wherein

the judgment step judges the ruled line candidate as unstable when the positional arrangement of the ruled line elements included in the ruled line candidate is not uniform.

10. The program according to claim 2, wherein

the judgment step judges the ruled line candidate as unstable when circumscribed rectangles of the ruled line elements constituting the ruled line candidate overlap circumscribed rectangles of black pixel link components other than the ruled line elements and having predetermined profiles.

11. The program according to claim 2, wherein

the judgment step extracts a plurality of ruled line candidates uniformly arranged in the vertical direction and a plurality of ruled line candidates uniformly arranged in the horizontal direction as a group of regular ruled line candidates and judges the ruled line candidates other than the group of regular ruled line candidates as unstable.

12. The program according to claim 11, wherein

the judgment step puts ruled line candidates having a starting point located within a predetermined range from a first y-coordinate, an ending point located within a predetermined range from a second y-coordinate and a same number of ruled line elements and ruled line candidates having a starting point located within a predetermined range from a first x-coordinate, an ending point located within a predetermined range from a second x-coordinate and a same number of ruled line elements into the group of regular ruled line candidates.

13. The program according to claim 2, wherein

the requirement determination step extracts a ruled line candidate contained in a predetermined region located in the vicinity of each ruled line candidate judged as unstable in the judgment step out of the ruled line candidates judged as stable in the judgment step and defines a value between the parameter value indicating the properties of the ruled line candidate and a threshold value of the first requirement as a threshold value of the second requirement.

14. A ruled line extracting apparatus that extracts a ruled line in an image of a document, comprising:

an extraction section that extracts a ruled line candidate from the image of a document according to a first requirement predefined to be met by figures of elements of the ruled lines;

a judgment section that judges if the ruled line candidate is stable or unstable according to a structural stability of the ruled line candidate extracted by the extraction section;

a requirement determination section that determines a second requirement to be met by the figures of the elements of the ruled line different from the first requirement according to the ruled line candidate judged as stable by the judgment section and the first requirement; and

a re-extraction section that re-extracts a ruled line candidate according to the second requirement determined by the requirement determination section.

15. The apparatus according to claim 14, wherein

the first requirement includes a parameter range showing the range of an element parameter that indicates properties of a graphic to be selected as a ruled line element of the ruled line and/or the range of a consolidation parameter that indicates properties of the graphic of the ruled line element to be consolidated as ruled line candidate.

16. The apparatus according to claim 15, wherein

the extraction section is adapted to select a graphic that makes the element parameter to be found within the corresponding parameter range from the document image and consolidate a plurality of ruled line elements that make the consolidation parameter to be found within the corresponding parameter range to generate the ruled line candidate.

17. The apparatus according to claim 16, wherein

the ruled line is a broken ruled line;

the parameter range of the element parameter is defined by a threshold value for sizes of circumscribed rectangles of black pixel link components; and

the extraction section being adapted to extract the black pixel link components from a document image expressed by a binary system and select the black pixel link components that make the element parameter to be found within the corresponding parameter range as ruled line elements.

18. The apparatus according to claim 16, wherein

the ruled line is a broken ruled line;

the parameter range of the consolidation parameter is defined by a threshold value for a gap separating two ruled line elements, a threshold value for a drift in the direction perpendicular to the direction of consolidating two ruled line elements and/or a threshold value for the number of consolidated ruled line elements; and

the extraction section being adapted to select a plurality of ruled line elements that make the consolidation parameter to be found within the parameter range as ruled line candidate.

19. The apparatus according to claim 16, wherein

the ruled line is a solid ruled line;

and

the extraction section is adapted to divide a document image expressed by a binary system into a plurality of regions, select regions that make the element parameter to be found within the parameter range out of the regions as ruled line elements and consolidate the ruled line elements to generate a ruled line candidate.

20. A ruled line extracting method of extracting a ruled line in an image of a document, the method comprising:

an extraction step that extracts a ruled line candidate from the image of a document according to a first requirement predefined to be met by figures of elements of the ruled lines;

a judgment step that judges if the ruled line candidate is stable or unstable according to the structural stability of the ruled line candidate extracted in the extraction step;

a requirement determination step that determines a second requirement to be met by the figures of the elements of the ruled line different from the first requirement according to the ruled line candidate judged as stable in the judgment step and the first requirement; and

a re-extraction step that re-extracts a ruled line candidate according to the second requirement determined in the requirement determination step.