实时图像匹配辅助无人机导航研究

Research of Unmanned Aerial Vehicle Navigation Aided by Real-Time Image Matching Technology

针对GPS/INS组合导航系统在抗干扰性和精度上不能完全满足无人机导航系统要求的问题,利用实时图像匹配辅助无人机导航的方法进行仿真研究。首先,对输入图像进行目标识别定位,为了满足系统的实时性要求,研究采用光电混合联合变换相关器进行实时图像匹配,仿真结果表明,目标的识别定位精度小于两个像素,能够为无人机提供精确的导航信息;再根据摄影测量原理计算导航信息中的位置和姿态角信息;最后,将匹配定位得到的导航信息通过Kalman滤波器与惯导系统解算的导航信息进行信息融合,仿真结果表明,滤波后导航信息的误差均获得了较明显的收敛,提高了无人机导航系统的抗干扰性和导航精度。

The anti-jamming and precision of GPS/INS integrated navigation system cannot meet the requirements of unmanned aerial vehicle （UAV） navigation system. To improve the anti-jamming and navigation precision of UAV, the UAV navigation aided by real-time image matching technology is researched. Firstly, the location of the target is obtained after target recognition. To meet the real-time requirements of the system, the image matching is carried out by using the joint transform correlator. Computer simulation shows that the location accuracy is better than 2 pixels, which can achieve a real-time image matching aided UAV navigation. Then, the position and attitude of the navigation information are calculated by the principle of photogrammetry. Finally, the navigation parameters obtained by image matching are used to integrate with the inertial navigation system data in an optimal filtering approach （ Kalman filter）. Computer simulation shows that both the anti-jamming and precision of the UAV navigation system are improved.