2d lidar and camera fusion for object detection and object distance measurement of ADAS using robotic operating system (ROS)

Abstract

The advanced driver assistance systems (ADAS) are one of the issues protecting people from a vehicle collision. A collision warning system is an essential part of ADAS to protect people from the dangers of accidents caused by fatigue, drowsiness, and other human errors. Multi-sensors have been widely used in ADAS for environment perception such as cameras, radar, and light detection and ranging (LiDAR). This work proposes that the relative orientation and translation between the two sensors must be considered in performing fusion. The researchers discuss the real-time collision warning system using 2D LiDAR and Camera sensors for environment perception and measure the distance (range) and angle of obstacles. In this paper, the researchers propose a fusion of two sensors consisting of a camera and 2D LiDAR to get the distance and angle of an obstacle in front of the vehicle implemented on Nvidia Jetson Nano using Robot Operating System (ROS). Hence, a calibration process between the camera and 2D LiDAR is required. After that, the integration and testing were carried out using static and dynamic scenarios in the relevant environment. The fusion process's experimental results between the camera and 2D LiDAR obtained an error rate of 0.197 meters. For better accuracy results on object detection and object distance measurement in the upcoming research, it is recommended to use the computational geometric transformation and projection approach.