Design and implementation of a robotic car to recognize traffic signs

This paper describes the design and implementation of a robotic car that can detect traffic signs in real environment and perform maneuvers accordingly. A webcam mounted on the car captures live images of the environment which are then processed using an onboard laptop where a simple template matching algorithm is used for the detection of these signs. Upon a successful match, an appropriate steering command is issued to the car. A single input single output fuzzy logic controller running on an 8bit AT89C52 microcontroller ensures the tracking of these steering commands. A number of experiments are performed in corridor environments and successful results are obtained.