A noise tolerant algorithm for the object recognition of warning system

In this communication, we discuss a noise tolerant traffic sign recognition algorithm which can be utilized in future vehicles to warn drivers that they are approaching traffic signs at an intersection. It is assumed that the vehicle is equipped with a video camera providing chromatic images of the navigational environment. The primary objective of this paper is to develop a noise tolerant color segmentation algorithm and a recognition algorithm which is tolerant to the rotation, position, and scale variations. The treatment of traffic signs by computer vision techniques has been fairly limited in the literature.