一种基于SLAM的无人机影像快速三维重建方法

A Fast 3D Reconstruction Method of SLAM-based UAV Image

无人机技术在电力通道巡检、选线等应用中发挥中越来越重要的作用。传统的基于影像的三维重建方法是以离线方式获取三维重建结果,效率不高。文章提出一种基于SLAM的无人机影像快速三维重建方法,通过视觉SLAM获取无人机序列影像的初始位姿信息,建立图像匹配优先度队列,有效减少无效的影像匹配,同时融合无人机GPS结果,实现快速的三维重建。实验表明,文章所提方法在效率上得到了显著提升,并且精度满足1:1000大比例尺测图要求。

UAV technology plays a more and more important role in the application of power channel inspection, line selection and so on. The traditional image-based 3 D reconstruction method is to obtain 3 D reconstruction results offline, and the efficiency is not high. In this paper, a fast 3 D reconstruction method of UAV image based on SLAM is proposed. the initial position and pose information of UAV sequence image is obtained by visual SLAM, and the priority queue of image matching is established, which can effectively reduce the invalid image matching. At the same time, the GPS results of UAV are fused to realize fast 3 D reconstruction.The experimental results show that the efficiency of the method proposed in this paper has been significantly improved, and the accuracy meets the requirements of 1:1000 large-scale mapping.