Abstract In this research we propose a real-time system to recognize blue traffic signs designating directions. This research is complementary to the previous work done on six annular red signs. The system consists of several processing steps: We firstly label the blue objects in each frame and segment them from the background. After that we try to verify if the segmented blue object is a sign candidate, and then we segment white objects within the blue object. Finally we classify the white objects by matching them to arrow patterns according to geometrical features, or reject them if no arrow pattern is matched. Classification is done using a decision tree. Processing time is about 110 ms/frame, and recognition rate is about 81%.
Real-Time Recognition of Blue Traffic Signs Designating Directions
International Journal of Intelligent Transportation Systems Research ; 8 , 2 ; 96-105
2010-03-24
10 pages
Article (Journal)
Electronic Resource
English
Blue traffic signs recognition , Arrow signs , Real-time recognition , Color segmentation Engineering , User Interfaces and Human Computer Interaction , Civil Engineering , Computer Imaging, Vision, Pattern Recognition and Graphics , Robotics and Automation , Automotive Engineering , Electrical Engineering
Real-time detection and recognition of road traffic signs
Online Contents | 2012
|Lane departure warning and real-time recognition of traffic signs
Tema Archive | 2009
|Lane Departure Warning and Real-time Recognition of Traffic Signs
British Library Conference Proceedings | 2009
|Traffic information delivery system designating direction and distance
European Patent Office | 2020
|