Title :
Recognition of traffic signs in color images
Author :
Gege Mo ; Aoki, Yuya
Author_Institution :
Graduate School of Information Science and Technology, Hokkaido University, Sapporo, Japan
Abstract :
In this paper, a method is proposed for using a camera installed in an automobile to detect and recognizing a traffic sign from a photographic image to prevent a driver from overlooking the traffic sign. In this study, "SI value" witch means saturation change with intensity, is proposed as a new color parameter. The shape of a candidate sign is determined by a symmetric-framework shape determining method, and a sign area is detected. The detection experiment was conducted on 500 images of four kinds of signs, blue, red, yellow and red-blue.
Keywords :
Automobiles; Cameras; Color; Image recognition; Information science; Pixel; Road accidents; Road vehicles; Shape; Vehicle driving;
Conference_Titel :
TENCON 2004. 2004 IEEE Region 10 Conference
Conference_Location :
Chiang Mai
Print_ISBN :
0-7803-8560-8
DOI :
10.1109/TENCON.2004.1414541