A LiDAR System Simulator Using Parallel Raytracing and Validated by Comparison with a Real Sensor

Abstract The advances in 3D imaging and Computer Vision are allowing for a massive acquisition of data, especially in the form of point clouds. However, scanning with these sensors can be costly and time consuming, reducing the efficiency of certain procedures, like deep learning dataset generation and forest canopy calculation. One trending solution is the creation of 3D imaging sensor simulators. This paper presents a simulator for light detection and ranging (LiDAR) systems with a parallel raytracing approach and a flexible scene creator. Finally, the simulator sensor is validated with a real LiDAR data.