Abstracts - faqs.org

Abstracts

Electronics

Search abstracts:
Abstracts » Electronics

On-line handwriting recognition

Article Abstract:

On-line handwriting recognition machines that recognize each character as it is being written are ideal for large-alphabet languages, like Japanese. Handwriting input using an on-line recognition method is necessary for input accuracy and speed, and it offers several advantages over off-line handwriting recognition. However, there are problems that impede high recognition accuracy without introducing handwriting constraints. For example, Kanji can be written with wide variations in the number and order of strokes and significant distortions of shape. In addition, to create a better man-machine interface, the writing of box-free character recognition is required. Nippon Denshin Denwa's research in this field over the past 15 years is described, and the man-machine interfaces made possible with on-line handwriting recognition examined.

Author: Wakahara, Toru, Murase, Hiroshi, Odaka, Kazumi
Publisher: Institute of Electrical and Electronics Engineers, Inc.
Publication Name: Proceedings of the IEEE
Subject: Electronics
ISSN: 0018-9219
Year: 1992
Research, Industrial research, Nippon Telephone and Telegraph Corp., Research and Development, User Interface, User-Friendliness, User interfaces (Computers), Handwriting, Kanji

User Contributions:

Comment about this article or add new information about this topic:

CAPTCHA


Document analysis - from pixels to contents

Article Abstract:

A conceptual framework is presented for solving the document analysis task. The task involves converting a document's pixel representation into an equivalent knowledge network representation holding the document's layout and content. The system contains several levels of abstraction. At the pixel level, elementary geometric objects are formed with regard to layout analysis and the definition of character objects; character recognition achieves the mapping from geometric object to character meaning in ASCII representation. On the next level, words are formed and verified through contextual processing. At the highest level of abstraction, modeled knowledge concerning complete documents and about the relation of their constituents to the application form are handled. The problems encountered at each level are discussed.

Author: Schurmann, Jurgen, Bartneck, Norbert, Bayer, Thomas, Franke, Jurgen, Mandler, Eberhard, Oberlander, Matthias
Publisher: Institute of Electrical and Electronics Engineers, Inc.
Publication Name: Proceedings of the IEEE
Subject: Electronics
ISSN: 0018-9219
Year: 1992
Pattern recognition (Computers), Word processing software, Modeling, Data modeling software, Layout, Pattern Recognition

User Contributions:

Comment about this article or add new information about this topic:

CAPTCHA



Subjects list: Computer science, Technical, Character Recognition
Similar abstracts:
  • Abstracts: On-board switching and processing. Communications switching - from operators to photonics
  • Abstracts: On the generation of random numbers from heavy-tailed distributions
  • Abstracts: On the computation of motion from sequences of images - a review. Imaging radar polarimetry: a review
  • Abstracts: Synthetic neural modeling: the "Darwin" series of recognition automata
  • Abstracts: Petri nets: properties, analysis and applications. Entropy nets: from decision trees to neural networks. Modern permanent magnets for applications in electro-technology
This website is not affiliated with document authors or copyright owners. This page is provided for informational purposes only. Unintentional errors are possible.
Some parts © 2025 Advameg, Inc.