无人机斜面贴近摄影测量的三维航线规划研究

Research on 3D route planning of UAV slope nap-of-the-object photogrammetry

针对倾斜摄影测量对与水平面存在较大夹角的斜面目标,难以采集到有效影像数据这一问题,本文提出基于主成分分析(principal component analysis,PCA)算法的斜面贴近摄影测量三维航线规划方法。首先,利用PCA算法求取无人机的飞行方向与摄影机主光轴方向;其次,结合约束条件得到的航迹点间隔,计算出航迹点三维坐标与无人机姿态;最后,生成无人机航线文件并导入地面控制端飞行作业,实现斜面目标高分辨率影像数据自动采集。结果表明:通过本方法得到的三维航线布局工整,采集到的影像分布均匀,其重建的三维模型精度高、纹理细腻、几何完整性好;相较于倾斜摄影测量得到的模型,在点位相对精度、模型完整性、纹理精细度上得到了显著提升。

The flight path planning methods in photogrammetry are mostly equal altitude flight or ground simulation flight.However,when facing highly variable and irregular ground,such as high slopes,dams,building facades,etc.,the target image data obtained using traditional route planning methods typically has missing data,especially at the bottom of the target.The three reconstructed models often include large areas of wrinkles and holes,which do not meet the requirements of refinement.Therefore,in view of the limitations of traditional flight paths,this paper proposes a slope nap-of-the-object photogrammetry 3D flight path planning method based on the PCA algorithm.The PCA algorithm is used to determine the UAV’s flight direction and the direction of camera’s main optical axis.Then,the interval of track points obtained from the constraints is used to calculate the track points’3D coordinates and the UAV attitude.Finally,in order to automatically acquire high-resolution image data of inclined targets,the UAV flight path file is generated and imported into the ground control terminal for flight operations.This effectively solves the problem of acquiring high-resolution image data of inclined targets.Experiments show that the 3D route layout obtained through this method is neat,the collected images are evenly distributed,and the reconstructed 3D model has high accuracy,fine texture,and good geometric integrity.Compared to the model obtained from oblique photogrammetry,it has significantly improved in terms of point-to-point relative accuracy,model integrity,and texture fineness.Experiments verify the feasibility of the slope nap-of-the-object photogrammetry route planning method presented in this paper.In the planning of flight paths for close-range photogrammetry of inclined targets,this method is simpler and more efficient than rotating the inclined plane,and also provides a new idea for 3D flight path planning of slope nap-of-the-object photogrammetry.The experiment shows that the texture of the model obtained by nap-of-the-object photogrammetry is fine and complete,and the relative accuracy of the internal point locations reaches 5.5 mm.In practical production,any complex object can be segmented into multiple slopes with small curvature and small surface undulations.For a single slope,a 3D flight path can be planned using the method proposed in this article.By combining the flight paths for each slope,it is possible to achieve slope nap-of-the-object photogrammetry 3D flight path planning for complex targets.Therefore,the method proposed in this paper has strong applicability and good application prospects.