Improving Driver Safety through Automated Traffic Sign Recognition Systems

This research presents a CNN-based automated traffic sign recognition system designed to enhance driver safety by accurately detecting and classifying road signs in real time. The proposed TS-CNN model is trained on the GTSRB dataset, incorporating techniques like HSV segmentation, shape-based feature extraction, and data augmentation to improve robustness under various conditions. With a classification accuracy of 95% and fast inference speed, the system aids in reducing accidents caused by missed or misunderstood traffic signs, making it suitable for real-world driver assistance applications.