A SIMPLIFIED TRAFFIC SIGN DETECTION AND RECOGNITION SYSTEM USING CONVOLUTIONAL NEURAL NETWORKS

The development in automotive intelligent technology in ADAS (Advanced Driver Assistance System) has made Traffic sign detection and recognition play an important role in expert systems. It instantly assists drivers or automatic driving systems in detecting and recognizing traffic signs effectively. This work proposed a method to design real-time traffic sign detection and recognition system in a real traffic situation. The images of the road scene were converted to grayscale images and then filtered the grayscale images with simplified color segmentation techniques, where the parameters were optimized. Convolution Neural networks and support vector machines (SVM) were employed for the detection and classification of the traffic signs. An accuracy of about 95.3% was achieved as a comparable performance with the state-of-the-art method. Keywords