Table recognition method and apparatus, and storage medium
First Claim
Patent Images
1. A table recognition apparatus comprising:
- analysis means for analyzing an area and position of each cell in a table based on a layout state of ruled lines and character strings contained in an image of the table to obtain a table structure;
table structure display means for displaying the table structure obtained by said analysis means to indicate the area of every cell and which cell each character string belongs to;
instruction means which receives a correction instruction input by a user, wherein the correction instruction is for instructing correction of an area of a cell or which cell a character string belongs to in the table structure displayed by said display means;
correction means for correcting the area and position of the cell based on the correction instruction received by said instruction means to obtain a corrected table structure; and
generation means for recognizing characters in each cell of the corrected table structure and generating table format data based on a recognition result and the corrected table structure.
1 Assignment
0 Petitions
Accused Products
Abstract
The area and position of each cell in a table are analyzed on the basis of the layout state of ruled lines and character strings contained in the image of the table to obtain a table structure. The obtained table structure is displayed. When a user instructs to correct the area of a cell in the displayed table structure, the area and position of the cell are corrected on the basis of the correction instruction to obtain a corrected table structure. After correction of the table structure, characters in each cell of the corrected table structure are recognized, and table format data is generated and output on the basis of the recognition result and table structure.
31 Citations
14 Claims
-
1. A table recognition apparatus comprising:
-
analysis means for analyzing an area and position of each cell in a table based on a layout state of ruled lines and character strings contained in an image of the table to obtain a table structure;
table structure display means for displaying the table structure obtained by said analysis means to indicate the area of every cell and which cell each character string belongs to;
instruction means which receives a correction instruction input by a user, wherein the correction instruction is for instructing correction of an area of a cell or which cell a character string belongs to in the table structure displayed by said display means;
correction means for correcting the area and position of the cell based on the correction instruction received by said instruction means to obtain a corrected table structure; and
generation means for recognizing characters in each cell of the corrected table structure and generating table format data based on a recognition result and the corrected table structure. - View Dependent Claims (2, 3, 4, 5, 6)
ruled line extraction means for extracting the ruled lines contained in the image of the table; and
cell area extraction means for detecting a rectangular area surrounding characters contained in the image of the table and merging rectangular regions separated by a distance having not more than a predetermined value to extract rectangular areas of cells, wherein the table structure is obtained based on the ruled lines and the rectangular areas of the cells, which are extracted by said ruled line extraction means and said cell area extraction means.
-
-
4. The apparatus according to claim 3, wherein
said analysis means determines boundaries of columns and rows of the table based on the ruled lines extracted by said ruled line extraction means and space portions between the rectangular areas of the cells extracted by said cell area extraction means. -
5. The apparatus according to claim 4, wherein
said table structure display means superposes, on the image of the table, boundary lines representing the rectangular areas of the cells and the boundaries obtained by said analysis means. -
6. The apparatus according to claim 4, wherein
said table structure display means displays character string rectangles, cell area rectangles, and boundary lines between the cells of the image of the table.
-
7. A table recognition method comprising:
-
analyzing an area and position of each cell in a table based on a layout state of ruled lines and character strings contained in an image of the table to obtain a table structure;
displaying the table structure obtained in said analyzing step to indicate an area of every cell and which cell each character string belongs to;
receiving a correction instruction input by a user, wherein the correction instruction is for instructing correction of an area of a cell or which cell a character string belongs to in the table structure displayed in said display step;
correcting the area and position of the cell based on the correction instruction received in said receiving step to obtain a corrected table structure; and
recognizing characters in each cell of the corrected table structure and generating table format data based on a recognition result and the corrected table structure. - View Dependent Claims (8, 9, 10, 11, 12)
extracting the ruled lines contained in the image of the table; and
detecting a rectangular area surrounding characters contained in the image of the table and merging rectangular regions separated by a distance having not more than a predetermined value to extract rectangular areas of cells, wherein the table structure is obtained based on the ruled lines and the rectangular areas of the cells, which are extracted in said ruled line extracting step and said cell area extracting step.
-
-
10. The method according to claim 9, wherein said analyzing step comprises determining boundaries of columns and rows of the table based on the ruled lines extracted in said ruled line extracting step and space portions between the rectangular areas of the cells enacted in said cell area extracting step.
-
11. The method according to claim 10, wherein said table structure displaying step comprises superposing, on the image of the table, boundary lines representing the rectangular areas of the cells and the boundaries obtained in said analyzing step.
-
12. The method according to claim 10, wherein said table structure displaying step comprises displaying character string rectangles, cell area rectangles, and boundary lines between the cells of the image of the table.
-
13. A storage medium storing a control program for generating table format data on the basis of image information of a table, the control program comprising:
-
code for an analysis step of analyzing an area and position of each cell in the table based on a layout state of ruled lines and character strings contained in an image of the table to obtain a table structure;
code for a table structure display step of displaying the table structure obtained in the analysis step to indicate the area of every cell and which cell each character string belongs to;
code for an instruction step of receiving a correction instruction input by a user, wherein the correction instruction is for instructing correction of an area of a cell or which cell a character string belongs to in the table structure displayed in the display step;
code for a correction step of correcting the area and position of the cell based on the correction instruction received in the instruction step to obtain a corrected table structure; and
code for a generation step of recognizing characters in each cell of the corrected table structure and generating the table format data based on a recognition result and the corrected table structure.
-
-
14. A table recognition apparatus comprising:
-
an analysis unit for analyzing an area and position of each cell in a table based on a layout state of ruled lines and character strings contained in an image of the table to obtain a table structure;
a table structure display unit for displaying the table structure obtained by said analysis unit to indicate the area of every cell and which cell each character string belongs to;
an instruction unit which receives a correction instruction input by a user, wherein the correction instruction is for instructing correction of an area of a cell or which cell a character string belongs to in the table structure displayed by said display unit;
a correction unit for correcting the area and position of the cell based on the correction instruction received by said instruction unit to obtain a corrected table structure; and
a generation unit for recognizing characters in each cell of the corrected table structure and generating table format data based on a recognition result and the corrected table structure.
-
Specification