Document image recognition (DIR) is a part of document image understanding (DIU) or intelligent document processing (IDP) system. The objectives of document image recognition is data extraction either textual or graphical data which exist within the document, or structural information such as document layout or document style which will resulting in exact reconstruction of document. Data extracted from this system will be used in further application forming document image understanding system or automatic document processing system. Document image recognition and understanding has been studied over three decades. Many commercial or free software are available, However these software primarily being spesific application. This paper will discuss on the development of a flexible framework of a DIR system which can be applied to many field of document image recognition applications. The paper will also discuss the classification of document complexity that are being used, methods used, and some application prototypes of a DIR system built with this framework.
Iping Supriana Suwardi, Peb Ruswono Aryan, Bugi Wibowo; Bandung Institute of Technology
Iping Supriana Suwardi, Peb Ruswono Aryan, Bugi Wibowo; Bandung Institute of Technology
