Traffic sign detection is an important computer vision task for intelligent transportation systems, advanced driver assistance systems, and autonomous vehicles. Manual identification of traffic signs is time-consuming and may become unreliable when signs appear at different scales, under varying illumination, or within complex road scenes. This chapter presents a deep learning based traffic sign detection system developed using the YOLOv8 object detection model. The proposed system processes road images and identifies traffic signs by generating bounding boxes, class labels, and confidence scores. The model was trained to recognize fifteen traffic-related classes, including Green Light, Red Light, Speed Limit 10, Speed Limit 20, Speed Limit 30, Speed Limit 40, Speed Limit 50, Speed Limit 60, Speed Limit 70, Speed Limit 80, Speed Limit 90, Speed Limit 100, Speed Limit 110, Speed Limit 120, and Stop. The experimental prediction set contained 638 images. Model performance was analyzed using confusion matrices, precision-recall curves, precision-confidence curves, recall-confidence curves, and F1-confidence curves. The precision-recall analysis produced an overall mAP@0.5 of 0.976, while the precision-confidence curve reached 1.00 at a confidence threshold of approximately 0.975. The F1-confidence curve reported a maximum overall F1 score of 0.96 at a confidence value of approximately 0.416. The results demonstrate that the trained YOLOv8 model can detect and classify the targeted traffic signs with high confidence. The system can support applications in road safety, traffic monitoring, driver assistance, and intelligent transportation systems, while future improvements can focus on real-time deployment, challenging environmental conditions, and edge-device implementation and practical deployment support.

Sunmathi K R, Srinivasan S K, Santhosh D, Regalya Sri M, Dr.Priya K,"Traffic Sign Detection using YOLOv8" International Journal of Advanced Multidisciplinary Research and Educational Development, V2(4): Page(445-457) July - August 2026. ISSN: 3107-6513. www.ijamred.com. Published by Scientific and Academic Research Publishing.