Table region identification method
AssigneeRicoh Company, Ltd.
InventorMichiyoshi Tachikawa
Life of this patent
Estimating the term…
Abstract
A method for identifying a table region within an input image data of an original obtained by an optical character reader. The table region identification method comprises the steps of extracting first sets of runs of successive black picture elements from each of the lines of the input image data, unifying adjacent runs among the first sets of the runs into first rectangles, unifying second sets of runs within a second rectangle having a size greater than a prescribed size among the thus unified first rectangles to extract ruled lines, the second sets of the runs having a length greater than a prescribed length, and finally making a decision on whether the second rectangle constitutes a table region or any other region depending on the number of the ruled lines counted.
Claims6
A method of locating and identifying a table region within an input image data of an original, said method comprising the steps of:
- reading the input image data of the original to store the input image data in an image data memory;
- extracting first sets of runs of successively read black picture elements from each line of the input image data stored in the image data memory to store the first sets of the runs in a run memory, each said first set of the runs stored in the run memory including coordinate values of a starting point and an end point of each said run;
- unifying adjacent runs among the first sets of the runs stored in the run memory into first rectangles, coordinate values of each of said first rectangles being stored in a rectangle memory;
- extracting a second rectangle rom among the first rectangles stored in the rectangle memory, said second rectangle having a width greater than a prescribed width and a height greater than a prescribed height;
- unifying second sets of runs of successive black picture elements from each line within said second rectangle to extract ruled lines, each of said second sets of the runs having a length greater than a prescribed length, coordinate values of a starting point and an end point of each said ruled line being stored in a run data buffer;
- counting the number of the ruled lines thus extracted from said second sets of the runs stored in the run data buffer within said second rectangle;
- and
- making a decision on whether the counted number of the ruled lines is greater than a prescribed number, thus making a determination that said second rectangle makes up the table region in the input image data stored in the image data memory;
- the determination that the second rectangle makes up the table region within the original being made based on the number of the ruled lines within the second rectangle thus extracted from the input image data of the original which is stored in the image data memory.
Dependencies are inferred from the claim text.