A access traffic sign detection tracking and recognition system for video frames using KNN algorithms

Abstract

Video-based traffic sign detectiontracking and recognition plays an important role in driving support system as well as in intelligent autonomous vehicles. This framework includes three parts like traffic sign detecting, target tracking and sign recognition. This paper contributes different methods for detecting and recognizing of traffic sign, so that drivers can be easily identify the type of signal in road sides. In this approach different methods like color segmentation and scale based intraframe fusion technics is proposed which includes the spatial-temperol constraints in videos, by fusing the different casings that have a place with an equivalent physical sign along to induce higher exactness.