Apparatus and method for area separation for image, having improved separation accuracy

• US 5,757,957 A
• Filed: 03/01/1995
• Issued: 05/26/1998
• Est. Priority Date: 11/29/1991
• Status: Expired due to Term

• Current Assignee: Ricoh Company Limited
• Original Assignee: Ricoh Company Limited
• Inventors: Tachikawa, Michiyoshi
• Primary Examiner(s): Razavi, Michael T.
• Assistant Examiner(s): CHANG, JON CARLTON
• Application Number:
  US08/396,585
• Time in Patent Office: 1,182 Days
• Field of Search: 358/462, 358/464, 358/453, 382/173, 382/176, 382/177, 382/178
• US Class Current: 382/176
• 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

Abstract

A reference-character size determination unit determines a reference-character size based on a distribution of sizes of particular continuous image parts, the continuous image parts constituting an image. An image part sorting unit sorts the particular continuous image parts into character image parts and other image parts based on the reference-character size. A rectangle determination unit determines a rectangle for each continuous image part, a size of the rectangle corresponding to a size of the corresponding continuous image part. The reference-character size determination means determines the reference-character size based on a distribution of sizes of the rectangles. The image part sorting means sorts the particular rectangles into rectangles for the character image parts and rectangles for the other image parts based on the reference-character size. The character image part integrating unit integrates the particular character image parts with each other when particular character image parts have a predetermined character relationship.

Claims

1. An area separation apparatus comprising:

a) rectangle determination means for determining a rectangle enclosing each continuous image part constituting an image, wherein a size of the rectangle depends on the size and the shape of the corresponding continuous image part;

b) reference-character size determination means for determining a reference-character size based on a distribution of sizes of particular rectangles;

c) image part sorting means for sorting the continuous image parts into character image parts and other image parts, based on said reference-character size, said image part sorting means includes means for sorting the particular rectangles into rectangles for the character image parts and rectangles for the other image parts, based on said reference-character size;

d) character image part integrating means for integrating particular ones of said rectangles with each other when said particular rectangles have a predetermined character relationship, said character image part integrating means including;

1) a line-formation unit for forming formed lines as a result of integrating the particular rectangles with each other;

2) a character area formation unit for forming a character area as a result of integrating formed lines from the line-formation unit with each other; and

3) means for varying an integrating threshold value which determines whether or not two of said rectangles are integrated, the varying means constituting means for reducing the integrating threshold value when a ruled line is present between the two rectangles; and

e) a ruled-line detecting unit for detecting the ruled line.

2. The area separation apparatus according to claim 1, wherein:

a predetermined image part revising relationship is a relationship between the ruled line and a particular character image part, in which relationship a distance between the ruled line and the particular character image part is within a predetermined ruled-line distance.

3. The area separation apparatus according to claim 2, wherein:

said predetermined image part revising relationship is such that the ruled line is located in a position in a direction in which characters are written so that said ruled line and said particular character image part look as if they overlap with each other when they are viewed in the direction in which the characters are written.

4. The area separation apparatus according to claim 3, wherein said character image part integrating means integrates a second rectangle with a first rectangle except when conditions 1) and 2) apply:

1) said character image part integrating means revises a predetermined mutual distance used for determination of whether or not the rectangles are integrated with each other when said first formed line is formed, and

2) an end of a second formed line is located in a position in the direction of said first formed line.

5. An area separation apparatus comprising:

a) rectangle determination means for determining a rectangle enclosing each continuous image part constituting an image, wherein a size of the rectangle depends on the size and the shape of the corresponding continuous image part;

b) reference-character size determination means for determining a reference-character size based on a distribution of sizes of particular rectangles;

c) image part sorting means for sorting the particular image parts into character image parts and other image parts, based on said reference-character size said image part sorting means includes means for sorting the particular rectangles into rectangles for the character image parts and rectangles for the other image parts, based upon said reference-character size;

d) character image part integrating means for integrating particular ones of said rectangles with each other when said particular rectangles have a predetermined character relationship, said character image part integrating means including;

1) a line-formation unit for forming formed lines as a result of integrating the particular rectangles with each other;

2) a character area formation unit for forming a character area as a result of integrating formed lines from the line-formation unit with each other; and

3) means for varying an integrating threshold value which determines whether or not two rectangles are integrated, the varying means constituting means for reducing the integrating threshold value when a ruled line is present between the two rectangles; and

e) a ruled-line detecting unit for detecting the ruled line;

wherein said character image part integrating means integrates a second character area with a first character area except when conditions

     1) and

     2) apply;

1) a ruled line perpendicular thereto is present, said first character area crossing said ruled line; and

2) an end of the second character area is located at a position beyond a corresponding end of the first character area and in the direction of said first character area.

6. An area separation apparatus comprising:

a) rectangle determination means for determining a rectangle enclosing each continuous image part constituting an image, wherein a size of the rectangle depends on the size and the shape of the corresponding continuous image part;

b) reference-character size determination means for determining a reference-character size based on a distribution of sizes of particular rectangles;

c) image part sorting means for sorting the particular continuous image parts into character image parts and other image parts, based on said reference-character size;

d) character image part integrating means for integrating particular ones of said rectangles with each other when said particular rectangles have a predetermined character relationship, said character image part integrating means including;

1) a line-formation unit for forming formed lines as a result of integrating the particular rectangles with each other;

2) a character area formation unit for forming a character area as a result of integrating formed lines from the line-formation unit with each other; and

3) means for varying an integrating threshold value which determines whether or not two of said rectangles are integrated, the varying means constituting means for reducing the integrating threshold value when a ruled line is present between the two of said rectangles; and

e) a character area integrating unit for integrating character areas from the character area formation unit; and

f) a ruled-line detecting unit for detecting the ruled line;

wherein said character area integrating unit integrates a second character area with a first character area except when conditions

     1) and

     2) apply;

1) the ruled line perpendicular thereto is present such that one of said formed lines would cross said ruled line; and

2) an end of the second character area and said ruled line are located at a position beyond a corresponding end of the first character area in said direction of said first character area.

7. An area separation apparatus comprising:

a) rectangle determination means for determining a rectangle enclosing each continuous image part constituting an image, wherein a size of the rectangle depends on the size and the shape of the corresponding continuous image part;

b) reference-character size determination means for determining a reference-character size based on a distribution of sizes of particular rectangles;

c) image part sorting means for sorting the continuous image parts into character image parts and other image parts, based on said reference-character size;

d) character image part integrating means for integrating particular ones of said rectangles with each other when said particular rectangles have a predetermined character relationship, said character image part integrating means including;

1) a line-formation unit for forming formed lines as a result of integrating the particular rectangles with each other;

2) a character area formation for forming a character area as a result of integrating the formed lines with each other, said character area formation unit includes means for integrating the formed lines when the formed lines have a predetermined line relationship which is based on a mutual distance between the formed lines measured along a direction perpendicular to a direction of the formed lines; and

3) means for varying integrating threshold value which determines whether or not two of said rectangles are integrated, the varying means constituting means for reducing the integrating threshold value when a ruled line is present between the two of said rectangles; and

e) a ruled-line detecting unit for detecting the ruled line;

wherein said character area formation unit includes;

means for revising said mutual distance into a revised mutual distance when a physical relationship between said ruled line and one of the formed lines is a predetermined line revising relationship corresponding to said physical relationship between said ruled line and the particular formed line.

8. The area separation apparatus according to claim 7, wherein:

said predetermined line revising relationship is a relationship between the ruled line and the particular formed line, in which relationship a distance between the ruled line and the particular formed line is within a predetermined line distance.

9. The area separation apparatus according to claim 8, wherein:

said predetermined line revising relationship is such that the ruled line is located in a direction in which said particular formed line has been extended and they look as if they overlap with each other when they are viewed in said direction of said particular formed line.

10. The area separation apparatus according to claim 9, further comprising:

a character area integrating unit;

wherein said character area integrating unit integrates a second character area with a first character area except when conditions

     1) and

     2) apply;

1) said character area formation unit revises said predetermined mutual distance used for determining whether or not the rectangles are integrated with each other when the formed line is formed, said formed line being included in the first character area, and

2) an end of the second character area and said ruled line are located at a position beyond a corresponding end of the first character area in said direction of the first character area.

11. An area separation apparatus comprising:

a) rectangle determination means for determining a rectangle for each continuous image part, wherein a size of the rectangle to be determined depends on the size and the shape of the corresponding continuous image part;

b) reference-character size determination means for determining a reference-character size based on a distribution of sizes of particular rectangles;

c) image part sorting means for sorting the continuous image parts into character image parts and other image parts, based on said reference-character size;

d) character image part integrating means for integrating particular ones of said rectangles enclosing the image parts with each other when said particular rectangles have a predetermined character relationship, said character image part integrating means further including;

1) means for varying an integrating threshold value which determines whether or not two rectangles are integrated, the varying means constituting means for reducing the integrating threshold value when a ruled line is present between the two rectangles; and

2) a border detecting unit for detecting a type of continuous image part constituting a border continuous image part, said border continuous image part being a continuous image part located at a position within a predetermined distance from a border of a large continuous image part of said continuous image part having predetermined threshold dimensions, said border continuous image part being a continuous image part having a predetermined shape resembling a ruled line, said image part integrating means failing to integrate the character image parts each located at mutually opposite sides with respect to said border continuous image part.

12. An area separation method comprising the steps of:

a) determining a rectangle for each continuous image part, wherein a size of the rectangle to be determined depends on the size and the shape of the corresponding continuous image part;

b) determining a reference-character size based on a distribution of sizes of particular rectangles, said continuous image parts constituting an image;

c) sorting the continuous image parts into character image parts and other image parts, based on said reference-character size, said image part sorting step b) including sorting of the particular rectangles into rectangles for the character image parts and rectangles for the other image parts, based on said reference-character size;

d) integrating particular ones of said rectangles enclosing the image parts with each other when said particular character rectangles have a predetermined character relationship, wherein said integrating step d) includes;

d1) forming formed lines as a result of integrating particular ones of said rectangles with each other; and

d2) forming a character area as a result of integrating said formed lines with each other;

wherein the integrating step also includes;

d3) varying an integrating threshold value which determines whether of not two of said rectangles are integrated, the varying step constituting a step of reducing the integrating threshold value when a ruled line is present between said two rectangles; and

e) detecting the ruled line.

13. The area separation method according to claim 12, wherein:

a predetermined image part revising relationship is a relationship between the ruled line and a particular character image part, in which relationship a distance between the ruled line and the particular character image part is within a predetermined ruled-line distance.

14. The area separation method according to claim 13, wherein:

said predetermined image part revising relationship is such that the ruled line is located in a position in a direction in which characters are written so that said ruled line and said particular character image part look as if they overlap with each other when they are viewed in the direction in which the characters are written.

15. The area separation method according to claim 14 wherein said step (d-2) of forming the character area fails to integrate a second formed line with a first formed line when said step (d) of integrating the rectangles revises a predetermined mutual distance used for determination whether or not the rectangles are integrated with each other when said first formed line is formed, and when an end of the second formed line is located in a position in the direction of said first formed line, said end of the second formed line being an end located in the direction of said second formed line.

16. An area separation method comprising the steps of:

a) determining a rectangle for each continuous image part, wherein a size of the rectangle to be determined depends on the size and the shape of the corresponding continuous image part;

b) determining a reference-character size based on a distribution of sizes of particular rectangles, said continuous image parts constituting an image;

c) sorting the continuous image parts into character image parts and other image parts, based on said reference-character size, said image part sorting step b) including sorting of the particular rectangles into rectangles for the character image parts and rectangles for the other image parts, based on said reference-character size;

d) integrating particular ones of said rectangles enclosing the image parts with each other when said particular character rectangles have a predetermined character relationship, wherein said integrating step d) includes;

d1) forming formed lines as a result of integrating particular ones of said rectangles with each other; and

d2) forming a character area as a result of integrating said formed lines with each other;

wherein the integrating step also includes;

d3) varying an integrating threshold value which determines whether or not two of said rectangles are integrated, the varying step constituting a step of reducing the integrating threshold value when a ruled line is present between said two rectangles; and

e) detecting the ruled line;

wherein said step d) of integrating the rectangles includes integrating a second character area with a first character area except when conditions A) and B) apply;

A) a perpendicular ruled line is present such that one of said formed lines would cross said ruled line; and

B) an end of the second character area and said ruled line are located at a position beyond a corresponding end of the first character area in said direction of said first character area.

17. An area separation method comprising the steps of:

a) determining a rectangle for each continuous image part, wherein a size of the rectangle to be determined depends on the size and the shape of the corresponding continuous image part;

b) determining a reference-character size based on a distribution of sizes of particular rectangles, said continuous image parts constituting an image;

c) sorting the continuous image parts into character image parts and other image parts, based on said reference-character size, said image part sorting step b) including sorting of the particular rectangles into rectangles for the character image parts and rectangles for the other image part, based on said reference-character size;

d) integrating particular ones of said rectangles enclosing the image parts with each other when said particular rectangles have a predetermined relationship, wherein said integrating step d) includes;

d1) forming formed lines as a result of integrating particular ones of said rectangles with each other; and

d2) forming a character area as a result of integrating said formed lines with each other;

wherein the integrating step also includes;

d3) varying an integrating threshold value which determines whether or not two of said rectangles are integrated, the varying step d3) constituting a step of reducing the integrating threshold value when a ruled line is present between the two rectangles;

e) integrating character areas formed in the step d2) of forming a character area; and

f) detecting the ruled line;

wherein said step e) of integrating the character area includes integrating a second one of the character area with a first character area except when conditions A) and B) apply;

A) a ruled line perpendicular thereto is present such that said first character area would cross said ruled line; and

B) an end of the second character area and said ruled line are located at a position beyond a corresponding end of said first character area in said direction of said first character area.

18. An area separation method comprising the steps of:

a) determining a rectangle for each continuous image part, wherein a size of the rectangle to be determined depends on the size and the shape of the corresponding continuous image part;

b) determining a reference-character size based on a distribution of sizes of particular rectangles;

c) sorting the particular continuous image parts into character image parts and other image parts, based on said reference-character size, said sorting step c) includes a step of sorting the particular rectangles into rectangles for the character image parts and rectangles for the other image parts, based upon said reference-character size;

d) integrating particular ones of said rectangles enclosing the image parts with each other when said particular rectangles have a predetermining character relationship, wherein said integrating step d) includes;

d1) forming formed lines as a result of integrating particular ones of said rectangles with each other; and

d2) forming a character area as a result of integrating said formed lines with each other, said step d2) including integrating the formed lines when the formed lines have a predetermined line relationship which is based on a mutual distance between the formed lines, said mutual distance being measured along a direction perpendicular to a direction along the formed lines;

d3) varying an integrating threshold value which determines whether or not two of said rectangles are integrated, the varying step constituting a step of reducing the integrating threshold value when a ruled line is present between the two rectangles; and

e) detecting the ruled line; and

wherein said step d2) of forming the character area includes;

revising said mutual distance into a revised mutual distance when a physical relationship between said ruled line and the particular formed line is a predetermined line revising relationship, corresponding to said physical relationship between said ruled line and the particular formed line.

19. The area separation method according to claim 18, wherein:

said predetermined line revising relationship is a relationship between the ruled line and the particular formed line, in which relationship a distance between the ruled line and the particular formed line is within a predetermined line distance.

20. The area separation method according to claim 19, wherein:

said predetermined line revising relationship is further a relationship between the ruled line and the particular formed line, in which relationship the ruled line is located in a direction in which said particular formed line has been extended and in which relationship they look as if they overlap with each other when they are viewed in said direction of said particular line.

21. The area separation method according to claim 20, further comprising an additional step of integrating the character areas, said additional step failing to include integrating a second character area with a first character area when said step of forming the character area includes revising said predetermined mutual distance used for determination whether or not the rectangles are integrated with each other when the formed line is formed, said formed line being included in the first character area, and when an end of the second character area and said ruled line are located at a position beyond a corresponding end of the first character area in said direction of the first character area.

22. An area separation method comprising the steps of:

a) determining a rectangle for each continuous image part, wherein a size of the rectangle to be determined depends on the size and the shape of the corresponding continuous image part;

b) determining a reference-character size based on a distribution of sizes of particular rectangles;

c) sorting the particular continuous image parts into character image parts and other image parts, based on said reference-character size; and

d) integrating particular ones of said rectangles enclosing the image parts with each other when said particular rectangles have a predetermined character relationship, wherein said integrating step also includes;

d'"'"') varying an integrating threshold value which determines whether or not two of said rectangles are integrated, the varying step constituting a step of reducing the integrating threshold value when a ruled line is present between the two rectangles;

wherein;

1) said step d) of integrating the rectangles includes detecting a type of continuous image part constituting a border continuous image part which is a continuous image part which is located in a position within a predetermined distance from a border of a large continuous image part of said continuous image part having predetermined threshold dimensions, and which has a predetermined shape resembling a ruled line;

but

2) said step d) of integrating the rectangles fails to include integrating the rectangles located at mutually opposite sides with respect to said border continuous image part.