Obstacle Detection in Autonomous Vehicles Using 3D LiDAR Point Cloud Data

Driving vehicles from one place to another safely and on time is not so easy. Situations like busy traffic and other health or weather conditions make it more complex, leading it to the development of the intelligent self-driving vehicles. Light detection and ranging (LiDAR) is the sensor that helps them detect the obstacles and avoid collisions. In this paper, a novel method is proposed for obstacle detection in autonomous vehicles. Open source 3D LiDAR point cloud data obtained when Velodyne HDL-64E LiDAR sensor is scanned in a real world environment is processed using Point Cloud Library (PCL). Random sample consensus (RANSAC) and K-Dimension tree algorithms are used for obstacle detection.