Implementation of an Internet of Vehicle System for High-Speed Traffic Collision Accident Detection with an Emergency Alert Mechanism

• Main Authors: SU, KE-YU, 蘇科毓
• Other Authors: CHANG, WAN-JUNG
• Format: Others
• Language: zh-TW
• Published: 2018
• Online Access: http://ndltd.ncl.edu.tw/handle/67e36s

碩士 === 南臺科技大學 === 電子工程系 === 106 === A research survey report pointed out that the rescue response time required for high-speed traffic accident collisions was one of the main factors affecting the road traffic accident mortality rate. Most traffic collision accidents could be got relevant assistances by drivers, passengers, or kindhearted persons. However, if a traffic accident occurs at night or in the suburbs, often driving or occupants lose consciousness due to violent impact, no one can immediately send a distress message to the outside world and affect the golden window the traffic collision accident. Therefore, this research implementation of an Internet of vehicle system for high-speed traffic collision accident detection with an emergency alert mechanism, mainly including a collision sensing module, an in-vehicle infotainment (IVI) platform, and cloud information platform. When possible traffic collision event is detected, the vehicle front image information will be uploaded to the cloud information platform through the IVI platform, and the accident image depth learning algorithm on the platform will be used to identify whether the image in front of the vehicle is a collision accident. If it is a collision accident, the cloud information platform can immediately issue an alarm message for a high-speed traffic collision accident. The experimental simulation results prove that the system proposed in this paper has extremely high detection accuracy of high-speed traffic collision accident. Therefore, when a high-speed traffic collision accident occurs, in addition to immediate detection, it can be immediately notified to shorten the gold rescue. Time will effectively improve the survival rate of the injured.