Human performance is prone to error in case of taking optimal decision on driving issues. It is due to the lack of concentration or because of some faulty characteristics of human nature. This is one of the core reasons for road accidents in our country. To diminish away this fact, autonomous vehicle system can be a profound solution. Also in modern technological aspects, it is higher seeking concept now. Addressing these events, we attempted to implement an autonomous vehicle system with the aid of computer vision and neural network based learning process. The system learns from image frames from a camera and real-time direction command corresponding to every frame. Then it moves autonomously by matching the learned frames with the current frames through neural network. It is also capable of detecting obstacle, stop and traffic signals and act accordingly.
A Supervised Learning Approach to An Unmanned Autonomous Vehicle
2019-05-01
3375794 byte
Conference paper
Electronic Resource
English