Forward collision mitigation system is a safety feature available on cars. This feature prevents collision caused by the lack of maintaining a safe distance between vehicles, as well as from drivers who lack concentration. However, not all vehicles have this feature. So in this research, an android-base application whose function and objectives are similar to the safety feature will be made. This application will use TensorFlow lite for the Custom Object Detector, SSD MobileNet V2 for the pre-trained model, and OpenCV library for the application framework. The research is done using the CRISP-DM methodology, as well as literature studies from previous research as a comparison. The output of this research is an application that can give a passive alert to the driver if the distance between vehicles is less than 3 m for cars, less than 2 m for motorcycles and less than 5 m for buses and trucks, giving a safe distance to do braking if needed.
Object Detection Application for a Forward Collision Early Warning System Using TensorFlow Lite on Android
Lect. Notes in Networks, Syst.
Congress on Intelligent Systems ; 2022 September 05, 2022 - September 06, 2022
19.05.2023
14 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
Reliability assessment method of forward collision early warning system
Europäisches Patentamt | 2023
|Vehicle forward anti-collision early warning system and method
Europäisches Patentamt | 2022
|