| Research Title |
Scanned Printed Thai Character Recognition Using the Bitmap-to-sequence Transformation |
| Date of Distribution |
14 January 2013 |
| Conference |
| Title of the Conference |
The 2013 International Conference on Embedded System and Intelligent Technology |
| Organiser |
Thai Embedded System Association, Embedded System R&D Group, Khon Kaen University and Faculty Engineering Khon Kaen University |
| Conference Place |
Asawann Hotel |
| Province/State |
Nong Khai |
| Conference Date |
13 January 2013 |
| To |
15 January 2013 |
| Proceeding Paper |
| Volume |
1 |
| Issue |
1 |
| Page |
156-159 |
| Editors/edition/publisher |
|
| Abstract |
Abstract— This paper presents a novel simple method for
recognition of scanned printed Thai characters using the bitmapto-
pixel sequence transformation which tracks pixels of
skeletonized character images resulting in signature graphs of the
characters. The proposed algorithm tracks pixel coordinates of
connected pixels, pixel by pixel, on each branch of the skeleton.
When the tracking process reaches a junction of branches, the
algorithm will use a predefined circular table to determine a
branch to be tracked next. Later, the dynamic time warping
method is then applied to classify the obtained signature graphs.
Empirical results tested using Thai Angsana UPC fonts of sizes
14, 16, 18 and 20 points with various font styles reveal the
average classification accuracy of 89 percents. |
| Author |
|
| Peer Review Status |
มีผู้ประเมินอิสระ |
| Level of Conference |
นานาชาติ |
| Type of Proceeding |
Full paper |
| Type of Presentation |
Oral |
| Part of thesis |
true |
| Presentation awarding |
false |
| Attach file |
|
| Citation |
0
|
|
|
Publication
Journal Publication
ไทย