Traffic congestion and accidents are emerging as significant social problems due to technological advancements and urban population growth. While computer vision is valuable for detecting traffic incidents, environmental and technological constraints often hinder its performance, resulting in inefficiencies. To improve traffic safety and efficiency, deep learning algorithms like Convolutional Neural Networks (CNNs) and object detection models such as You Only Look Once (YOLO) and Faster Region-based Convolutional Neural Network (R-CNN) are increasingly utilized. A system using object tracking techniques calculates and displays the speed of moving vehicles in real-time using models like YOLO-NAS (You Only Look Once Neural Architecture Search) and Deep SORT (Deep Simple Online Realtime Tracking). The model, implemented with Python and OpenCV, analyses video streams to detect vehicles and calculate their speeds. The data processed can act as input for a future real-time traffic monitoring and control system.