Title
Distinguishing pattern-types in printed documents
Abstract
Reading machines and other devices have helped a lot in integrating blind persons into society. Simultaneously the progress and widespread use of desktop-publishing systems lead to documents that are hard to recognise for current reading machines. Nevertheless, access to printed documents is required by most blind persons. In this article we show a possible structure of an improved and flexible system for document recognition. It will be capable of processing any printed document, for instance a letter or a musical score as well as diagrams or line drawings. Special emphasis is put on correct and efficient distinction of categories of documents. Requirements for distinction are shown and an suggestion for an implementation is presented. It will perform this classification in a fast and flexible way. Also, the suggested method can automatically as well as interactively classify the type of a document. Afterwards the contents of the document can be displayed in an appropriate style. The blind users also have the possibility to interactively define new classes of documents if required. The prototype of this system is currently being implemented at our institute.
Year
DOI
Venue
1994
10.1007/3-540-58476-5_128
ICCHP
Keywords
Field
DocType
printed document,distinguishing pattern-types
Output device,Information retrieval,Computer science,Document recognition,Optical character recognition,Feature extraction,Blind persons,Line drawings
Conference
ISBN
Citations 
PageRank 
3-540-58476-5
1
0.48
References 
Authors
2
2
Name
Order
Citations
PageRank
Georg Lokowandt110.48
Waltraud Schweikhardt2367.59