Methods and devices for extracting text from documents

A text extraction device for extracting text from a document, the text extraction device comprises:

• a processor;
• and
• a memory communicatively coupled to the processor, wherein the memory stores instructions, which, on execution by the processor, causes the processor to:
• perform a layout analysis on the document to identify a plurality of regions within a plurality of pages in the document;
• identify a table region from within the plurality of regions based on homogeneity between a plurality of textual lines in a page from the plurality of pages, wherein the homogeneity is computed based on a plurality of preselected textual parameters associated with the plurality of textual lines, wherein identifying the table region further comprises:
• determine for each textual line in a page column within the page, values for the plurality of preselected textual parameters comprising at least one of pixel length, number of words, total pixel space between adjacent words, or number of characters;
• compute, for each textual line in the page column, a variance of value of at least one of the plurality of preselected textual parameter from an associated average parameter value determined for all textual lines within the page column;
• determine, for each textual line in the page column, an average variance based on the variance computed for each of the at least one of the plurality of preselected textual parameters;
• and
• compute, for each textual line in the page column, a covariance based on a difference between the average variance of each textual line and an associated contiguous textual line within the page column;
• identify at least two rows and at least two columns within the table region;
• identify a plurality of cells within the table region based on the at least two rows and the at least two columns;
• and
• extract text from each of the plurality of cells.

A non-transitory computer-readable storage medium comprising a set of executable instructions stored thereon that, when executed by one or more processors, cause the processors to:

• perform a layout analysis on the document to identify a plurality of regions within a plurality of pages in the document;
• identify a table region from within the plurality of regions based on homogeneity between a plurality of textual lines in a page from the plurality of pages, wherein the homogeneity is computed based on a plurality of preselected textual parameters associated with the plurality of textual lines, wherein identifying the table region further comprises:
• determine for each textual line in a page column within the page, values for the plurality of preselected textual parameters comprising at least one of pixel length, number of words, total pixel space between adjacent words, or number of characters;
• compute, for each textual line in the page column, a variance of value of at least one of the plurality of preselected textual parameter from an associated average parameter value determined for all textual lines within the page column;
• determine, for each textual line in the page column, an average variance based on the variance computed for each of the at least one of the plurality of preselected textual parameters;
• and
• compute, for each textual line in the page column, a covariance based on a difference between the average variance of each textual line and an associated contiguous textual line within the page column;
• identify at least two rows and at least two columns within the table region;
• identify a plurality of cells within the table region based on the at least two rows and the at least two columns;
• and
• extract text from each of the plurality of cells.