抽象的な

COGNITIVE WAY OF CLASSIFYING DOCUMENTS: A PRACTITIONER APPROACH

Muheet Ahmed Butt, Majid Zaman

Documents are the sources of data which result in information and knowledge. Processing documents to extract their contents in an automated fashion is an essential task in all types of organizations for varied applications. The classification of documents being processed is required for their efficient recognition as it reduces the number of searches and also reduces the chances of error at different stages during the process. Therefore, in this proposed research a robust classification mechanism for document images based on the layout structure of its different elements which results in cognition based recognition is presented. The document image is considered to contain text only or text as well as tables and images. The classification is based on a scheme of preserving the structure of the layout of a document image. The algorithms are based on the spatial relationships existing among the visual components present in the document.