1SPRS, Vol.34, Part 2W2, “Dynamic and Multi-Dimensional GIS", Bangkok, May 23-25, 2001 
246 
The recognition result was provided very accurately because the 
gray-level image processing is simple and fast to calculate the 
template matching. Moreover the sign images do not have the 
pattern variation. However, still, this step is not a real-time 
processing depending on the hardware. 
Example of Sign Positioning by a new algorithm 
The result was calculated (see table 2) and compared the 
accuracy with the conventional method (see the figure 9). 
Table 2 each sign position by the indirect measurement 
Northing(Meter) 
Easting(Meter) 
Sign No.l 
1554294.8 
647194.1 
Sign No.2 
1554266.4 
647046.8 
Sign No.3 
1554272.9 
646760.6 
Sign No.4 
1554289.3 
646318.5 
Sign No.5 
1554231.7 
646156 
Sign No.6 
1554136.7 
646136.8 
Figure 11 the accuracy of the new method compared with conventional method 
The accuracy of the new method compared with 
conventional method 
After sensor was calibrated, the sign position was repeated and 
recalculated with the new algorithm. The best result is a one- 
meter difference of the position in the Sign No.3 compared with 
the high-accurate sign position from the RTK-GPS conventional 
method. 
5 Conclusion 
This research is the beginning of the Thai Road Sign Recognition 
and Location. The algorithm has been developed simply but the 
accurate result was obtained specially for the Road Sign 
Recognition. However there are still limitations to study for the 
real application. Therefore the further study should considered 
with the Single GPS and a variety of study area. 
Acknowledgement 
Specially thank you for Space Technology Application and 
Research Program, Asian Institute of Technology to give me very 
skilledful education. Thank you for Asian Center of Remote 
Sensing Research to support the RTK-GPS sensor. 
6 Reference 
Chen X., Murai S., 1998, Automated Mapping: An Introduction of 
Map Image Interpretation, Asian Institute of Technology, 87-91 
Dana, P. H, 1997, Global Positioning System Overview, web site 
http://utxass.edu/depts/qrg/gcraft/notes/gps/gps/gps.html, 
Department of Geography, University of Taxes at Austin. 
DGPS InQuest: Demonstration Software for GPS, 
URL:http://www.quailtveng.co.uk/dqps/dqps1.html 
Dinesh, M, 1998, Procedural Development for Level Surveying 
Using Real Time Kinematic GPS, M.Eng. Thesis, School of 
Environment, Resources and Development, The Asian Institute 
of Technology, Bangkok, Thailand 
Geomatics Canada, Geodetic Surveying division, Accuracy 
Standards for Positioning 
Donald H., M.Paluline Baker, 1994, Computer Graphic:Affine 
Transformation,208