基于无人机视频的公路可量测三维重建方法

Road 3D reconstruction and measurement method based on unmanned aerial vehicle video

针对目前无人机视频信息在公路建设和管理中未能得到充分利用的现状,提出了一种基于低空无人机视频流影像的可量测三维重建方法。该方法首先是对无人机摄像机CCD传感器进行检校,然后直接从无人机获取的视频中自动提取关键帧影像,再根据提取关键帧时间信息对全球导航卫星系统信息进行内插和同步得到关键帧影像拍摄位置信息;最后,通过多视角多视图三维重建技术,生成彩色密集点云数据并进行三维量测分析,实现基于无人机视频图像的量测和分析。通过试验分析,利用该方法可以根据无人机获取的动态视频信息,快速进行高精度三维场景重建和量测分析,可满足公路建设和管理中远程勘测调查的精度、大范围和快速性的要求。

Aiming to the video information of UAV not being fully utilized in highway construction and management,a measurable three-dimensional reconstruction method based on low-altitude UAV video stream image is proposed.Firstly,the CCD sensor of UAV camera is calibrated,and then the key frame image is automatically extracted from the video captured by UAV.Then the key frame image position information is obtained by interpolating and synchronizing the GNSS information based on the time information of extracted key frame.Finally,the color dense point cloud is generated by multi-view 3D reconstruction technology.The data are measured and analyzed in three dimensions to realize the measurement and analysis based on UAV video images.According to the dynamic video information acquired by UAV,the method can quickly reconstruct the high-precision three-dimensional scene and measurements.It can meet the requirements of the accuracy,large-scale and rapidity of long-distance survey in the highway construction and management.