TSDRDL: Traffic Sign Detection and Recognition using Deep Learning Techniques

The recognition and detection of traffic-signs are crucial components of expert systems, including automated driving and traffic assistance systems. These systems rely on quickly detecting and identifying traffic-signs to ensure safe and efficient driving. This research proposes a novel method for accurately detecting and recognizing traffic-signs in real-world situations. The approach involves pre-processing images by converting them to grayscale and filtering them. The traffic-signs are then classified into subclasses using a convolutional neural network (CNN) with filtered input images and the same parameters as the detection stage. The CNN architecture parameters are optimized to achieve the highest recognition. The Experimental-results demonstrate the CNN architecture outperforms prior studies regarding accuracy. Overall, this research offers a promising approach to improving the reliability and efficiency of automated driving and traffic assistance systems.