Optimized YOLO model to control traffic flow dynamically by analysing traffic density and conditions

As we know, there are a lot of rush and traffics are there in the township or so-called urban cities due to the increasing level of traffic congestion issues as the number of automobiles grows exponentially, a greater number of economic losses, environmental dilapidation, and reduced eminence of life. This novel work specifies a general machine learning model that focusses to optimize the flow of traffic by dynamically verifying traffic density and conditions in real-time. This whole work demonstrates YOLO object detection model to perceive and classify vehicles with a great level of accuracy i.e. 97.8% rate for different vehicle classes. The proposed solution employs computer vision, deep learning, and IoT technologies to develop a more robust traffic signal control system that specifies signal timings based on real-time traffic density instead of pre-programmed schedules. The model used Random Forest Regression and Support Vector Regression methods to predict traffic patterns, which is further facilitating proactive signal control. Implementation results show a 30% decrease in wait time during peak traffic and a 40% decrease in response time for emergency vehicles.