The number of vehicles on the road at a time is increasing rapidly these days. This generates a heavy traffic queue on the roads hence a proper traffic management system is needed to manage those traffic queues and most of the current traffic queues are handled by manual traffic management systems. That will require more human interaction and can lead to an increase in the stress level of the controllers, low efficiency in responding to emergency situations and will lead to more human errors. Considering the current traffic management scenarios, this research helps to enhance the road safety by developing a smart junction that has the embedded systems of junction coordination and synchronization by controlling traffic lights, emergency vehicle detection and prioritization, rule violation incidents and vehicle accidents identification and a safe pedestrian crossing system controlled based on dynamic factors. This study uses a Convolutional Neural Network (CNN) Network Architecture with Internet of Things (IOT) devices to automate the system. The significance of this research is that this will help to enhance overall road safety.