Forward Collision Avoidance System in a Self-driving Remote Controlled Car: An Application of Machine Intelligence

Abstract

Abstract: The paper discusses the experimental vision-based forward collision avoidance system (FORCAS), an autonomous bot car based on machine learning using Raspberry Pi as a processing platform. Using HD camera images as a primary input, Raspberry Pi combines processes and predict path using convolutional neural network and distance estimation models. It distinguishes the stationary objects on a drivable road surface and estimates their distances from a single camera’s view. Classification algorithms run on real-time images controls the steering of the car. On the basis of a detected object, it decides to avoid, stop or run over it.

Keywords: Forward collision avoidance system (FORCAS), Raspberry Pi, Tensorflow, remote controlled car

Cite this Article: Zankhana H. Shah, Tirth Patel, Vatsal Patel, Adhunik Rajput. Forward Collision Avoidance System in a Self-driving Remote Controlled Car: An Application of Machine Intelligence. Journal of Artificial Intelligence Research & Advances. 2019; 6(3): 13–18p.