Advanced Driver Assistance System for Autonomous Vehicles Using Deep Neural Network
Rahul Barnwal, Rishabh Kumar Srivastava, R Vimalathithan, S Kala · 2022
Connected vehicle technology has been on demand in recent years for improving the efficiency and safety of vehicles. Advancements in connectivity and computation will enable vehicles to be used as development platforms which can generate big data and perform feed forward computations, thereby enhancing the growth of transportation and economy. In this paper we propose an Internet of Things (IoT) enabled system that is capable of real time tracking of various parameters of a vehicle which includes vehicle's speed, RPM (Revolutions Per Minute), battery level and fuel level that can help in tracking of vehicle's health. An on-board camera has also been included in the proposed system that captures multimedia in real time and analyses the object detected. This camera based Advanced Driver Assistance System (ADAS) can help in avoiding accidents up to a greater extent. ADAS in this work uses DetectNet Deep Neural Network (DNN) for detecting the objects. We implement and analyze the performance of an ADAS system on an Nvidia Jetson Nano Developer Kit and the results are reported.