Traffic Sign Recognition System for Autonomous Vehicles using Deep Learning

The architecture of a fully autonomous car must be incorporated with a traffic sign recognition system. The Traffic Sign Recognition (TSR) consists of two components: detection and classification. The proposed study, which focuses on identifying these signals, is based on LISA dataset, which is the largest publicly accessible collection of images of traffic signs in the world. The performance of both aggregate channel features-based and integral channel feature-based detection approaches has reached its quality. In the proposed study, the performance of Convolutional Neural Networks (CNN), aggregate channel characteristics, and integral channel features are evaluated. The proposed study investigates the detection performance of CNN by tuning the convolutional layers, max-pool layers, and linear layers. The effectiveness of the proposed detection model is tested using the PASCAL measure, a typical statistic for this system.