Online Handwriting Recognition System
Online Handwriting Recognition system (HWRS) is a software use to convert handwritten character into text format. This project highlights the development of online handwriting recognition system using Microsoft Visual Basic 6.0, Support Vector Machine (SVM) and, VBTablet 2.0 to recognize the inpu...
Saved in:
| Main Author: | |
|---|---|
| Format: | Monograph |
| Published: |
UTeM
2009
|
| Subjects: | |
| Online Access: | http://library.utem.edu.my:8000/elmu/index.jsp?module=webopac-d&action=fullDisplayRetriever.jsp&szMaterialNo=0000054316 http://library.utem.edu.my:8000/elmu/index.jsp?module=webopac-d&action=fullDisplayRetriever.jsp&szMaterialNo=0000054316 http://eprints.utem.edu.my/976/1/ONLINE_HANDWRITING_RECOGNITION_SYSTEM_VOON_CHOI_PENG_TK7882.P3.V66_2009_-_24_pages.pdf http://eprints.utem.edu.my/976/2/ONLINE_HANDWRITING_RECOGNITION_SYSTEM_VOON_CHOI_PENG_TK7882.P3.V66_2009.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Online Handwriting Recognition system (HWRS) is a software use to convert
handwritten character into text format. This project highlights the development of
online handwriting recognition system using Microsoft Visual Basic 6.0, Support
Vector Machine (SVM) and, VBTablet 2.0 to recognize the input character by
matching the prototypes. This may take place by writing directly on to a digitizing
tablet by using stylus which is connected to the Universal Serial Bus (USB) port of a
computer. Owing to the fact that each individual has its own way of presenting
his/her handwriting on tablet, there is a certain level of complexity like the way of
holding the stylus, the strokes use in the writing and the amount of time and pressure
put on tablet which are involved in this recognition system. The general on-line
recognition procedures are preprocessing, features extraction, coarse classification,
detail matching and postprocessing. In this project the classifier used is the Support
Vector Machine (SVM) and a new extractor is proposed consisting of time and
pressure data, which provides of several more meaningful features of the handwritten
character. The experiment results show that the SVM trained using features from
time and pressure data makes the time and pressure technique interesting and
competitive since the features has certain recognition advantages such as a convex
objective function with very fast training algorithms. |
|---|