Abstract:
Traffic congestion is a serious problem, especially in urban areas where it continues to
deteriorate, making real-time road traffic density monitoring vital for efficient signal
control and traffic management. Various reasons contribute to congestion, such as
limited road capacity, unregulated demand, and extended red-light waits. While the first
two causes are partly connected, the constant nature of traffic signal delays, independent
of traffic circumstances, underscores the necessity for modeling and improving traffic
management to handle rising demand. In recent years, image processing and
surveillance technologies have become important to traffic control, delivering real-time
passenger information, ramp metering, and updates. Traffic density estimation may be
performed via image processing. This project intends to utilize live picture feeds from
cameras at traffic crossings to determine real-time traffic density and implement a
Signal Switching Algorithm that dynamically adjusts traffic light timings based on
vehicle density. This algorithm analyzes vehicle data to set optimal green signal
durations, ensuring that heavily trafficked lanes receive more time, while lanes with low
traffic do not experience unnecessary green lights. By adjusting signals in real-time, the
algorithm aims to minimize road congestion, reduce accidents, provide safer travel, and
cut fuel consumption and waiting times. Additionally, it will create valuable data for
future road design and research. In future phases, synchronizing several traffic signals
may further minimize congestion and enhance the free flow of traffic. Unlike previous
systems that depend on electrical sensors embedded in the pavement, this system
employs cameras mounted beside traffic signals to collect image sequences. Image
processing offers a more efficient approach for regulating traffic light status changes,
reducing congestion by preventing unnecessary green lights on empty lanes. It is also
more reliable in detecting vehicle presence by utilizing real traffic images,
demonstrating greater practicality and effectiveness compared to systems that detect
vehicles based on their metal content.
