ABSTRACT
The usage of vehicles has increased rapidly, and so the road accidents caused due to vehicles. Accidents generally occur due to various reasons like drowsiness of drivers, drink and drive. Unfortunately, accidents occurring on highways or any isolated areas do not get real-time rescues, and consequently, tolls caused due to accidents are increasing. This paper deals with providing a basic avoidance of accidents by checking the drowsiness of the driver and whether the driver is drunk or not. If somehow the accident cannot be avoided, then immediate detection of the accident will occur based on the sound of the collision, the most modern and unique approach. As sound is being used as one major parameter for accident detection, it gives us an accurate and quick response based on real-time data. Thus, it increases the detection efficiency of the system. The employment of fuzzy logic in accident detection and all the machine learning algorithms used for accident sound detection has enhanced detection performance. As soon as an accident is detected, the proposed system provides a smart rescue, irrespective of the presence or absence of an internet facility.
