基于多源数据融合的雅丹体三维建模

3D modeling of Yadang based on multi-source data fusion

为了解决倾斜摄影时由于遮挡问题导致的雅丹体建模容易产生漏洞的问题,采用地面拍摄像机、无人机以及三维激光扫描仪等设备采集多源空间数据,结合近景摄影测量、倾斜摄影测量和激光扫描技术,研究不同场景雅丹体建模方法,并对建模精度和效果进行了分析。实验结果表明,基于绕飞和近景影像,辅以激光点云数据精细化建模,相较于大场景倾斜建模,平面精度提高40%,高程精度提高16%,且表面纹理更加精细,为研究雅丹地貌变化和地质遗迹保护提供了高精度参考数据。

In order to improve the vulnerability during Yardang modeling caused by occlusion in oblique photography,multi-source spatial data was collected using equipment such as ground cameras,drones,and threedimensional(3D)laser scanners.By combining close-range photogrammetry,oblique photogrammetry,and laser scanning technology,Yardang modeling methods for different scenes were studied,and the modeling accuracy and effect were analyzed.The experimental results show that based on the flyby and close-range images,fine modeling is conducted by using laser point cloud data.Compared to large-scale oblique modeling,the plane accuracy is improved by 40%;the elevation accuracy is improved by 16%,and the surface texture is more refined.This can provide high-precision reference data for studying the changes in Yardang landforms and the protection of geological relics.