STRUCTURAL DECOMPOSITION IN HANDWRITING

The invention relates to a method for processing lists in handwriting (IN), comprising: initially classifying each of a plurality of text lines (LN) as a distinct text item (TI) which is not part of a list; and a classification process comprising a pattern detection in each text line (LN) for classi...

Full description

Saved in:
Bibliographic Details
Main Authors VERGNE, Julien, LORIANT, Nicolas
Format Patent
LanguageEnglish
French
German
Published 11.08.2021
Subjects
CALCULATING
COMPUTING
COUNTING
HANDLING RECORD CARRIERS
PHYSICS
PRESENTATION OF DATA
RECOGNITION OF DATA
RECORD CARRIERS
Online AccessGet full text

Cover

More Information
Summary:The invention relates to a method for processing lists in handwriting (IN), comprising: initially classifying each of a plurality of text lines (LN) as a distinct text item (TI) which is not part of a list; and a classification process comprising a pattern detection in each text line (LN) for classifying each text line starting with a predetermined list symbol (BT) as a distinct list item (LI) which is part of a list; determining an item indentation (22) of each text item (TI) with respect to a reference position (30) and determining for each list item (LI) a text indentation (24) representing the indentation of text comprised in said list item; and a merging step for merging, as part of a same text item (TI), or as part of a same list item (LI), if predefined conditions are met. A text structure data model may then be generated based on a result of the merging process, thereby defining each text line (LN) as part of either a text item (TI) or a list item (LI).
Bibliography:Application Number: EP20200305120