Abstract:
Numerous nations on the planet are confronting the issue at traffic light convergence that
causes mishaps between emergency vehicles and the other open vehicle. The quick
reaction of the crisis administrations, for example, ambulances or fire administration
vehicles has gotten a difficult circumstance nowadays. Some of the time the rescue
vehicle stalls out in rush hour gridlock and those minutes can cost people life. There are
existing systems to manage traffic light controls for emergency vehicles based on image
processing, Radio Frequency, and IR technology. But the current framework gave an
extraordinary opportunity to the emergency vehicles to release even non-emergency
circumstance time. So, in this project, we proposed an “AI-based traffic light control
system for emergency vehicles” that has allowed emergency vehicles to leave only in
actual emergencies. which can get the maximum benefits and save many lives. The main
objective of our research is identifying the emergency vehicles at intersections and doing
the more accurate AI-based traffic light control system to release them when stuck in
traffic jams to identify vehicles, we developed and trained object recognition models by
using image processing techniques especially for the ambulance, fire truck, and VIP
vehicles. The system identifies every object from the video, emergency vehicles were
considered as specifications to differentiate emergency vehicles from other vehicles. we
have designed the sound identification model to identify the siren sounds, here we have
trained varies siren sounds, our system which gain sound as an input from a microphone
and our system trained to filter noises to identify the emergency vehicle’s sirens sound
and combine both Sound & Image identification process when both conditions are true,
the system changes the red signal to green or extend the green signal duration by detect
the siren sound and emergency vehicles, and release the emergency vehicles path/way in
an emergency.
Keywords: Artificial intelligent, Emergency vehicle, Image processing, Machine learning,
Neural networks, Siren sound, Sound analysis, Traffic lights
