Document image processing method and apparatus

• US 20120045129A1
• Filed: 05/18/2011
• Published: 02/23/2012
• Est. Priority Date: 08/17/2010
• Status: Active Grant

• Current Assignee: PFU Limited (Fujitsu Limited)
• Original Assignee: Fujitsu Limited
• Inventors: Sun, Jun, Hotta, Yoshinobu, Yu, Hao, Naoi, Satoshi
• Granted Patent: US 8,345,977 B2
• Time in Patent Office: Days
• US Class Current: 382/176
• CPC Class Codes: G06V 30/1478 of characters or characters lines
• CPC Class Codes: G06V 30/413 Classification of content, e.g. text, photographs or tables
• CPC Class Codes: G06V 30/414 Extracting the geometrical structure, e.g. layout tree; Block segmentation, e.g. bounding boxes for graphics or text

Abstract

A method for processing a document image includes: performing horizontal and vertical text line extraction on the document image; providing an overlapping matrix, a value of an element of the overlapping matrix indicating an overlapping relation between horizontal and vertical text lines; merging the overlapping matrix in the vertical and horizontal direction; determining one or more text overlapping regions in the document image, based on the values of the elements of the merged overlapping matrix; counting the total number of strokes or pixel points in the horizontal and vertical text lines, respectively, within one of the one or more text overlapping regions; and determining an orientation of the text overlapping region is horizontal if the total number of strokes or pixel points in the horizontal text lines is larger than that in the vertical text lines, otherwise, determining the orientation is vertical.

Claims

1. A method for processing a document image, comprising:

performing horizontal text line extraction on the document image, to obtain horizontal text lines, the number of rows of the horizontal text lines being represented by Nh;

performing vertical text line extraction on the document image, to obtain vertical text lines, the number of columns of the vertical text lines being represented by Nv;

providing an overlapping matrix represented by MO with Nh rows and Nv columns, a value of an element represented by MO(i, j) of the ith row and the jth column of the overlapping matrix MO indicating an overlapping relation between the ith row of horizontal text lines and the jth column of vertical text lines, where 1≦

i≦

Nh and 1≦

j≦

Nv;

merging the overlapping matrix MO in the vertical direction, so that a value of an element of the overlapping matrix MO indicating an overlapping relation between a column of vertical text lines and each of a plurality of rows of horizontal text lines is set as a same value if the column of vertical text lines overlaps with the plurality of rows of horizontal text lines simultaneously;

merging the overlapping matrix MO in the horizontal direction, so that a value of an element of the overlapping matrix MO indicating an overlapping relation between a row of horizontal text lines and each of a plurality of columns of vertical text lines is set as a same value if the row of horizontal text lines overlaps with the plurality of columns of vertical text lines simultaneously;

determining one or more text overlapping regions in the document image, based on the values of the elements of the merged overlapping matrix MO;

counting the total number of strokes or pixel points in the horizontal text lines and in the vertical text lines, respectively, within one of the one or more text overlapping regions; and

determining an orientation of the one of the one or more text overlapping regions is a horizontal orientation if the total number of strokes or pixel points in the horizontal text lines is larger than that in the vertical text lines, otherwise, determining the orientation of the one of the one or more text overlapping regions is a vertical orientation.

2. The method according to claim 1, further comprising:

before determining one or more text overlapping regions in the document image, repeating the merging of the overlapping matrix MO in the vertical direction and in the horizontal direction until the values of the elements of the overlapping matrix MO do not change any more.

3. The method according to claim 1, wherein

MO(i, j)=0 indicates that there is no overlapping between the ith row of horizontal text lines and the jth column of vertical text lines;

MO(i, j)=i indicates that there is overlapping between the ith row of horizontal text lines and the jth column of vertical text lines;

a value of an element of the overlapping matrix MO indicating an overlapping relation between a column of vertical text lines and each of a plurality of rows of horizontal text lines is set as a minimum value thereof if the column of vertical text lines overlaps with the plurality of rows of horizontal text lines simultaneously; and

a value of an element of the overlapping matrix MO indicating an overlapping relation between a row of horizontal text lines and each of a plurality of columns of vertical text lines is set as a minimum value thereof if the row of horizontal text lines overlaps with the plurality of columns of vertical text lines simultaneously.

4. The method according to claim 1, further comprising:

determining the document image is in a horizontal orientation if it is determined that an orientation of each of the one or more text overlapping regions is the horizontal orientation;

determining the document image is in a vertical orientation if it is determined that the orientation of each of the one or more text overlapping regions is the vertical orientation; and

determining the document image is in a hybrid orientation if it is determined that the orientations of the one or more text overlapping regions include both the horizontal orientation and the vertical orientation.

5. An apparatus for processing a document image, comprising:

a horizontal text line extraction unit adapted to perform horizontal text line extraction on the document image, to obtain horizontal text lines, the number of rows of the horizontal text lines being represented by Nh;

a vertical text line extraction unit adapted to perform vertical text line extraction on the document image, to obtain vertical text lines, the number of columns of the vertical text lines being represented by Nv;

an overlapping matrix providing unit adapted to provide an overlapping matrix represented by MO with Nh rows and Nv columns, a value of an element represented by MO(i, j) of the ith row and the jth column of the overlapping matrix MO indicating an overlapping relation between the ith row of horizontal text lines and the jth column of vertical text lines, where 1≦

i≦

Nh and 1≦

j≦

Nv;

a vertical merging unit adapted to merge the overlapping matrix MO in the vertical direction, so that a value of an element of the overlapping matrix MO indicating an overlapping relation between a column of vertical text lines and each of a plurality of rows of horizontal text lines is set as a same value if the column of vertical text lines overlaps with the plurality of rows of horizontal text lines simultaneously;

a horizontal merging unit adapted to merge the overlapping matrix MO in the horizontal direction, so that a value of an element of the overlapping matrix MO indicating an overlapping relation between a row of horizontal text lines and each of a plurality of columns of vertical text lines is set as a same value if the row of horizontal text lines overlaps with the plurality of columns of vertical text lines simultaneously;

a text overlapping region determining unit adapted to determine one or more text overlapping regions in the document image, based on the values of the elements of the overlapping matrix MO merged by the vertical merging unit and the horizontal merging unit;

a counting unit adapted to count the total number of strokes or pixel points in the horizontal text lines and in the vertical text lines, respectively, within one of the one or more text overlapping regions determined by the text overlapping region determining unit; and

a text orientation determining unit adapted to determine an orientation of the one of the one or more text overlapping regions is a horizontal orientation if the total number of strokes or pixel points in the horizontal text lines counted by the counting unit is larger than that in the vertical text lines, otherwise, to determine the orientation of the one of the one or more text overlapping regions is a vertical orientation.

6. The apparatus according to claim 5, wherein, before the text overlapping region determining unit determines one or more text overlapping regions in the document image, the vertical merging unit and the horizontal merging unit repeat the merging of the overlapping matrix MO in the vertical direction and in the horizontal direction until the values of the elements of the overlapping matrix MO do not change any more.

7. The apparatus according to claim 5, wherein

MO(i, j)=0 indicates that there is no overlapping between the ith row of horizontal text lines and the jth column of vertical text lines;

MO(i, j)=i indicates that there is overlapping between the ith row of horizontal text lines and the jth column of vertical text lines;

the vertical merging unit sets a value of an element of the overlapping matrix MO indicating an overlapping relation between a column of vertical text lines and each of a plurality of rows of horizontal text lines as a minimum value thereof if the column of vertical text lines overlaps with the plurality of rows of horizontal text lines simultaneously; and

the horizontal merging unit sets a value of an element of the overlapping matrix MO indicating an overlapping relation between a row of horizontal text lines and each of a plurality of columns of vertical text lines as a minimum value thereof if the row of horizontal text lines overlaps with the plurality of columns of vertical text lines simultaneously.

8. The apparatus according to claim 5, further comprising an image orientation determining unit adapted to:

determine the document image is in a horizontal orientation if the text orientation determining unit determines that an orientation of each of the one or more text overlapping regions is the horizontal orientation;

determine the document image is in a vertical orientation if the text orientation determining unit determines that the orientation of each of the one or more text overlapping regions is the vertical orientation; and

determine the document image is in a hybrid orientation if the text orientation determining unit determines that the orientations of the one or more text overlapping regions include both the horizontal orientation and the vertical orientation.