Key points are not available for this paper at this time.
The integration of three-dimensional 3D mapping is crucial for applications in autonomous vehicles and environmental simulations. This paper presents the development of a real-time mapping system using GPS-IMU RTK, coupled with the Velodyne UltraPuck LiDAR scanner, installed on vehicles. The system aims to generate 3D maps utilizing various devices, such as long-range LiDAR sensors, to collect distance data across a full 360-degree horizontal and vertical field of view. This results in a Point Cloud dataset, complemented by GPS-IMU data providing coordinates, motion direction, and orientation information. The processing software, developed in C#, orchestrates the fusion of Point Cloud and GPS-IMU data to create accurate 3D maps. Coordination of data in the nanosecond range ensures precise alignment between LiDAR Sensor and GPS-IMU. The experiment demonstrates an achieved accuracy of 2.3 centimeters in the horizontal dimension and 2 centimeters in the vertical dimension. Algorithms play a vital role in this process, requiring a profound understanding of electronics and mathematical expertise. The Point Cloud dataset is transformed into 3D maps, referencing the Universal Transverse Mercator coordinate system (UTM). This allows for future integration of additional Point Cloud datasets onto existing maps. The research focuses on the development of a real-time 3D mapping system, showcasing results in a web browser through WebGL, developed using JavaScript and C#. The processing algorithms involve mathematical calculations to produce 3D maps with customizable radii and viewing angles. Additionally, the Point Cloud data is stored in the LAS file format, and data from LiDAR Sensor (PCAP) and vehicle motion (GPS-IMU) is recorded for further analysis and future improvements.
Noppapetthongkam et al. (Wed,) studied this question.