Positioning and orientation system (POS) based on UAV platform is integrated with differential GNSS high-precision positioning technology and INS navigation technology to provide position and attitude reference for aerial remote sensing based on direct geo-reference. It is the key to obtain high-quality remote sensing imaging. Aiming at the low-precision IMU adopted by the light and small UAV, the error model is researched and analyzed, and the 24th-order Kalman filter is established. Finally, the GNSS/INS loose combination mode is used to obtain high-precision POS data. The measured data shows that the POS data obtained under this error model can be combined with the radar data of the UAV to meet the laser point cloud accuracy requirements and effectively reduce the cost of the airborne laser scanning system.