Implementation of a Low-Cost Mini-UAV Laser Scanning System

Mini-UAV laser scanning system can be applied to the high precision earth observation, which is an active area in the field of mobile mapping. However, due to the limitation of payload and battery consumption of a mini-UAV (maximum payload is less than 5 kg), and the high cost of the sensors, a tradeoff must be made between cost, weight, and accuracy when designing of the mini-UAV laser scanning system. To realize the high precision and low-cost mobile mapping, this paper designs a low-cost mini-UAV laser scanning system:Luojia Kylin Cloud. This system contains several low-cost sensors, including MEMS (micro electro mechanical system) based IMU (inertial measurement unit), global shutter camera, wide angle lens and 16-line laser scanner. Firstly, this paper proposes an IMU aided bundle adjustment to improve the accuracy of the low-cost MEMS based IMU. Secondly, this paper proposes a boresight self-calibration algorithm for the laser scanner based on the consistence of the depth map generated by MVS(multi-view stereo) and projection of the laser measurement. At last, the laser point clouds are generated by using the estimated states a boresight parameters. To evaluate the accuracy of Luojia Kylin Cloud laser scanning system, study area in Wuhan University is selected for point cloud collection, and a lot of ground check points are measured. The result shows that the average error of the check points is 17.8 cm, which demonstrates the high accuracy and robustness of the proposed system.