公开(公告)号：MY177773A

公开(公告)日：2020-09-23

申请号：MYPI2015702102

申请日：2015-06-22

Applicant: MIMOS BERHAD

Inventor： WONG CHIN WEE ,  MOHAMAD HILMI MOHAMAD BAKHARI ,  LEE SUET HWA

Abstract: The present invention relates to an improved method and system for processing a scanned image of a document of any types and formats accurately and correctly while retrieving a document content therefrom. The present invention creates a document template simultaneously, on the fly, when reading the scanned image of the document without having the same pre-registered. The method preferably comprises determining size and shape of the document based on the scanned image (100) by an image processing unit (200); performing named-entity recognition and optical character recognition (101) by a content recognition unit (201); identifying a field region indicative of location of the document content on the scanned image (102) by a field region detection unit (202); creating a document template simulating the field region thereof, the size and the shape of the document (103) by a template processing unit (203); and performing template matching between the document template and existing templates stored in a repository (104) by the template processing unit (203).