Use of 3D laser radar for navigation of unmanned aerial and ground vehicles in urban and indoor environments

Abstract

This paper discusses the integration of Inertial measurements with measurements from a three-dimensional (3D) imaging sensor for position and attitude determination of unmanned aerial vehicles (UAV) and autonomous ground vehicles (AGV) in urban or indoor environments. To enable operation of UAVs and AGVs at any time in any environment a Precision Navigation, Attitude, and Time (PNAT) capability is required that is robust and not solely dependent on the Global Positioning System (GPS). In urban and indoor environments a GPS position capability may not only be unavailable due to shadowing, significant signal attenuation or multipath, but also due to intentional denial or deception. Although deep integration of GPS and Inertial Measurement Unit (IMU) data may prove to be a viable solution an alternative method is being discussed in this paper.The alternative solution is based on 3D imaging sensor technologies such as Flash Ladar (Laser Radar). Flash Ladar technology consists of a modulated laser emitter coupled with a focal plane array detector and the required optics. Like a conventional camera this sensor creates an "image" of the environment, but producing a 2D image where each pixel has associated intensity vales the flash Ladar generates an image where each pixel has an associated range and intensity value. Integration of flash Ladar with the attitude from the IMU allows creation of a 3-D scene. Current low-cost Flash Ladar technology is capable of greater than 100 x 100 pixel resolution with 5 mm depth resolution at a 30 Hz frame rate.The proposed algorithm first converts the 3D imaging sensor measurements to a point cloud of the 3D, next, significant environmental features such as planar features (walls), line features or point features (corners) are extracted and associated from one 3D imaging sensor frame to the next. Finally, characteristics of these features such as the normal or direction vectors are used to compute the platform position and attitude changes. These "delta" position and attitudes are then used calibrate the IMU. Note, that the IMU is not only required to form the point cloud of the environment expressed in the navigation frame, but also to perform association of the features from one flash Ladar frame to the next. This paper will discuss the performance of the proposed 3D imaging sensor feature extraction, position change estimator and attitude change estimator using both simulator data and data collected from a moving platform in an indoor environment. The former consists of data from a simulated IMU and flash Ladar installed on an aerial vehicle for various trajectories through an urban environment. The latter consists of measurements from a CSEM Swissranger 3D imaging sensor and a MicroStrain low-cost IMU. Data was collected on a manually operated aerial vehicle inside the Ohio University School of Electrical Engineering and Computer Science building.

Accurate 3-D Position and Orientation Method for Indoor Mobile Robot Navigation Based on Photoelectric Scanning

“…Thus, this location method is based on precise scanning angle measurement that is different from laser radar in principle [24]. Assuming that the moment when the receiver capture the synchronized laser signal is t 0 and that when it captures the two scanning signals are t 1 and t 2 , respectively, as the rotating speed is a constant value ω, then the scanning angle θ 1 of laser plane 1 from the initial position to the position when it passes through the receiver can be expressed as…”

3D Multi-Copter Navigation and Mapping Using GPS, Inertial, and LiDAR

“…Features could be extracted from the resulting 3D point cloud and used for 3D pose estimation. For example, Horn and Uijt de Haag describe 3D imager‐based methods that extract planar features from point cloud data and use these features to estimate the 6DOF pose either with or without an IMU. In order to use a 2D laser scanner for 6DOF pose estimation, it can be turned into a 3D sensor by using a motor to rotate the scanner or by using a gimbaled mirror mechanism .…”

Integration of 3D and 2D imaging data for assured navigation in unknown environments

“…Integration with secondary sensors is, therefore, required to mitigate this by periodically "resetting" the inertial errors. The navigation method proposed in this thesis, extends the solution discussed in [2], by adding the data from the 2D images into the solution. Combining the data of multiple types of sensors as well as considering multiple feature types should not only improve the accuracy of the position and attitude estimates, but also add integrity, continuity, and availability to the solution.…”