无人机倾斜摄影冗余数据删除算法

Redundant data deletion algorithm for UAV tilt photography

多镜头倾斜摄影测量技术同时从多个角度对地表成像,当无人机在临近测区边界飞行工作时,部分镜头拍摄的图像超出了测区范围,因此多镜头倾斜摄影测量技术通常存在大量冗余图像,冗余数据的存在导致后期的空三解算以及实景三维建模效率大大降低。针对这一问题,以5镜头倾斜相机为例,结合POS数据、相机参数、飞行参数等建立拍摄瞬间成像模型,快速解算每一张图像和测区范围之间的空间位置关系,从而删除冗余数据。实验证明,删除冗余数据,效率获得极大提升。

Multi-lens tilt photogrammetry technology images the surface from multiple angles at the same time.When the UAV flies near the boundary of the survey area,the images captured by some lenses are beyond the scope of the survey area.Therefore,there are usually a large number of redundant images in multi-lens tilt photogrammetry technology.The existence of redundant data greatly reduces the efficiency of aerial triangulation and real 3D modeling.To solve this problem,this paper takes a five lens tilt camera as an example,combines POS data,camera parameters,flight parameters and so on to establish the instantaneous imaging model,and quickly calculates the spatial position relationship between each image and the measuring area,so as to delete the redundant data.The experiments show that this method can greatly improve the processing efficiency by deleting redundant data.