针对大角度倾斜成像航空相机拍摄距离远, 激光测距设备作用距离有限的问题, 提出了一种不依赖距离测量设备的直接对地目标定位算法。依据载机POS测量的载机位置、姿态信息以及航空相机中位置编码器测量的框架角位置信息, 利用齐次坐标变换的方法求解目标在大地坐标系下的指向, 再利用地球椭球模型和数字高程模型确定目标点的经纬度信息。采用蒙特卡洛法仿真分析载机位置姿态测量误差及相机框架角位置误差对视轴指向精度的影响, 相比于仅采用地球椭球模型的目标定位算法, 该算法有效降低了地形起伏对目标定位影响, 在目标区域地形起伏标准差大于10 m时, 大角度倾斜成像的定位精度明显提高。采用飞行试验数据验证了该目标定位算法的有效性, 在飞行高度18 000 m拍摄框架横滚角小于63°时, 目标定位圆概率误差小于70 m, 可满足工程实际需要。

Aiming at problems that operating range of laser ranging finder limits shooting distance of aerial camera with large inclined angles, a directly target geo-location algorithm independent of ranging equipment was proposed. Based on aircraft positions and attitude information measured by airborne Position and Orientation System (POS), as well as outer and inner gimbal angles from the encoder in the aerial camera, target to sensor vector in Earth-Centered Earth-Fixed (ECEF) coordinate was solved by homogeneous coordinate transformation. Then according to ellipsoidal earth model and global Digital Elevation Model(DEM), the longitude and latitude of the target were determined. The Monte Carlo method was employed for simulated analysis on influences of measurement errors of aircraft position and altitude as well as encoder on pointing accuracy of Line of Sight (LOS). Compared with the geo-location algorithm which only based on ellipsoidal earth model, the algorithm effectively reduced the influence of uneven terrain on target location. When the standard deviation of uneven terrain in the target area is over 10 m, the positioning accuracy of the wide-angle inclined imaging is improved significantly. The flight test proves the feasibility of the target geo-location algorithm, in which the Circular Error Probability(CEP) of the target geo-location is smaller than 70 m when the flight height is 18 000 m and the roll angle of the photography inclination angle is less than 63°. Consequently the proposed algorithm can meet actual demands of the project.