On-line recognition of digits and alphabets

Pao Yue-kong Library Electronic Theses Database

On-line recognition of digits and alphabets

 

Author: Tsang, Chi-hang Peter
Title: On-line recognition of digits and alphabets
Degree: M.Sc.
Year: 2000
Subject: Optical pattern recognition
Writing -- Data processing
Imaging processing
Hong Kong Polytechnic University -- Dissertations
Department: Multi-disciplinary Studies
Dept. of Electronic and Information Engineering
Pages: vi, 81 leaves : ill. ; 30 cm
Language: English
InnoPac Record: http://library.polyu.edu.hk/record=b1527672
URI: http://theses.lib.polyu.edu.hk/handle/200/241
Abstract: Handwriting is a natural and fast way to input words to computers. Computer recognition of on-line handwriting offers a new way of improving the human-computer interface. The objective of this project is to study and implement efficient algorithms for recognising handwriting digits and alphabets. The system includes handwriting character capturing, feature extraction, off-line/on-line neural network training, and recognition of handwriting characters. Features used for recognition can be categorised into two groups: off-line and on-line features. On-line recognition makes use of pen trajectory data. Also spatially overlapping characters do not pose a serious segmentation problem as they are still separable by using the on-line information. On the other hand, on-line recognition should have a reliable and robust recognition unit. This can be achieved by using both on-line and off-line features, which are discussed in this dissertation. A multi-layer feedforward neural network is trained to recognise 62 handwriting characters (10 digits, 26 lowercase alphabets, and 26 uppercase alphabets) using the error back-propagation algorithm with a training data set of five samples for each character. Different weight vectors are trained for different writers. Experimental results show that the system can achieve a correct rate of about 89% for user-dependent recognition.

Files in this item

Files Size Format
b15276727.pdf 3.529Mb PDF
Copyright Undertaking
As a bona fide Library user, I declare that:
  1. I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
  2. I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
  3. I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.

     

Quick Search

Browse

More Information