Autonomous UAV technology is limited in its ability to land safely at distinct airfields that have not been precisely surveyed and where GPS is unavailable. In this paper, we present a multi-sensor system for the automatic landing of fixed-wing UAV. The system is composed of a high precision aircraft controller, a range finder (Lidar) and a vision module used for detection and tracking of runways. The estimation of the position of the fixed-wing UAV is by using Lidar and performs gliding till flaring. Then, a federated extended Kalman filter (EKF) structure is costumed and utilizes the solutions of the IMU, GPS and Lidar as independent measurements to estimate the position of the vehicle. The framework can be used to integrate the vision solutions and enables the estimation to be smooth and robust landing. For taxiing, the neural network is used such that from live video stream from the camera trains the UAV to land precisely along the runway.

中文翻译：

---

固定翼无人机激光雷达辅助自主着陆和视觉滑行

自主无人机技术在未经过精确测量和 GPS 不可用的不同机场安全着陆的能力受到限制。在本文中，我们提出了一种用于固定翼无人机自动着陆的多传感器系统。该系统由高精度飞机控制器、测距仪（激光雷达）和用于跑道检测和跟踪的视觉模块组成。固定翼无人机的位置估计是利用激光雷达进行滑翔直至爆发。然后，采用联合扩展卡尔曼滤波器 (EKF) 结构并利用 IMU、GPS 和激光雷达的解决方案作为独立测量来估计车辆的位置。该框架可用于集成视觉解决方案，并使估计能够平稳而稳健地落地。对于出租车，