One Real Time Detection and Positioning Method for Ground Targets by Unmanned Aerial Vehicles based on GNSS, MEMS, and Machine Vision Fusion

Ground target point positioning is an important function of unmanned aerial vehicle battlefield reconnaissance. Target point positioning requires high real-time and positioning accuracy. The objective of the paper is the real-time target positioning method of the unmanned aerial vehicles, which adopts the multiple information fusion methods. And the correction methods for coordinate errors in the method.A real-time using multiple information fusion method is proposed in the paper, the correction method for reducing coordinate errors is also proposed.The actual measurement results show that based on the above method, the target positioning accuracy has been significantly improved compared to traditional aerial target positioning methods. The oblique target positioning error can be controlled within 5 meters, and the vertical target positioning error is basically close to the GNSS error. This method will greatly improve the real-time performance and target positioning accuracy of unmanned aerial vehicle reconnaissance targets.