Signature mark detection

A system for detection of signature marks in a document, comprising:

• a candidate selector configured for selecting candidate text objects in an ordered set of document pages;
• a sequence detector configured for identifying sequences of elements detected in the candidate text objects, each detected element of the sequence occurring on a different page of the document, the sequence having a numbering pattern including an incremental part and optionally a fixed part, the sequence identification allowing for missing elements between two detected elements of the sequence;
• a model generator configured for generating models of the sequences identified by the sequence detector, each model storing the numbering pattern of the respective sequence, an increment, which is computed based on the distance between pages on which consecutive elements of the sequence are identified, a valid sequence having an increment of greater than 1, and a first page, which corresponds to the page of the document on which the sequence starts, the model generator combining sequences for which the respective models are equivalent;
• and
• a sequence validator configured for validating the sequences with the model, including at least one of:
• deleting elements which do not comply with the model, and
• searching for additional elements of the sequence not yet identified by considering only the content of pages which comply with the model.

A method for detection of signature marks in a document, comprising:

• receiving an ordered set of document pages into computer memory, the pages having been generated by scanning a multi-page hardcopy document;
• providing for identifying sequences of elements in the pages, each sequence having a numbering pattern, each detected element of a respective sequence, apart from the first, having an incremental part which is at least one higher than the incremental part of the previous element of the sequence detected on a previous page of the document, each detected element of the sequence using the same character set, and wherein up to a maximum number of missing elements between two detected elements of the sequence are permitted;
• providing for generating a respective model of the sequences identified, each model including:
• the numbering pattern of the sequence,
• an increment, which is computed based on the distance between the page on which the element is identified and the page on which the next element of the sequence is identified, for a plurality of the elements of the sequence, a valid sequence having an increment of greater than 1, and
• a first page, which corresponds to the page of the document on which the sequence starts, the first page being computed using the increment and page numbers for a plurality of the elements of the sequence;
• validating the sequence with the model;
• and
• for a valid sequence, identifying elements of the sequence in the pages of the document as signature marks.