Ref

Accident Detections

Limitations

Performance Measures

Simulators

[1]

Ultrasonic detector.

Limitation of processors on various devices.

Works effectively in a variety of street scenarios as well as in a wide range of natural conditions.

Infrared sensors (IR sensors), smart phones, airbags, and mobile applications.

[9]

Footage from security cameras placed on roads and highways that would suggest the possibility of a vehicle accident.

The method cannot identify vehicles that are so severely damaged that none of the vehicle parts under consideration are present.

Offers encouraging results when evaluated on two separate datasets of damaged cars, which vary in terms of the image’s quality, its proximity to the camera, and the amount of items in the image.

Damaged Cars Dataset-1 (DCD-1) and Damaged Cars Dataset-2 (DCD-2).

[10]

IoT system model-GMM and CART.

Requires constant Internet access.

Improved in terms of recall and precision.

G-force, rollover, fall-off, rollover, fire, accident location, and speed.

[11]

Intelligent traffic light system.

GSM messaging’s queue-based system’s delay.

Fully automated.

GSM messaging.

[6]

Accelerometer in car alarm application.

Internet stuttering and lost GSM during an accident.

Vibration sensor with increased sensitivity and accuracy.

GSM texting. GPS and an MCU ATmega 8A.

[12]

Accelerometer for micro-electromechanical systems with an ultrasonic sensor.

Not used in all cases.

Take immediate action and act quickly to save the victim’s life.

Many protocols for connectivity, including cellular, LoRaWAN, and NB-IoT.