Autonomous Car Using Deep Learning and Image Processing Technology

Manasa; Ramya G.; Shravan  Bembalagi; Prashanth S.; Charanraj B. R.

Abstract

This study showcases a mini version of autonomous cars where it includes both autopilot mode and manual control. Autopilot mode is implemented using deep learning and image processing techniques, such as CNN, Haar cascade model and mobile app for manual control. IoT plays an important role. The Raspberry pi, pi camera and Arduino uno work as the main processors. The high-resolution pi camera, which is of 8 MP, will provide real time information in the form of photos as input to raspberry pi which is mainly responsible for controlling the car based on its input, which in turn results in obstacle detection, traffic sign detection, stop sign detection and lane detection by using deep learning techniques. Along with these features, the car is able to overtake the other car if it comes across a lane with proper indicators.

Abstract

Keywords