多源数据融合实景三维建模关键技术研究

Study on Key Technologies of Real 3D Modeling Based on Multi Source Data Fusion

摘要:

本文以某市政府大楼为研究对象，获取车载移动测量及倾斜摄影测量数据，进行了精度分析、多源数据融合，基于融合成果进行实景三维模型重建。通过比对单独基于倾斜摄影的实景三维模型效果，分析了多源数据的融合精度及模型的改善效果。创新性地采用了基于车载移动测量成果的特征矢量要素自动提取技术，结合国产北斗基站数据进行倾斜航摄差分解算，改进了实景三维建模技术的广泛应用。

Abstract:

In this paper, the improvement effect of multisource data fusion on oblique real 3D reconstruction has been studied. Taking Tengzhou city government building as the research object, the data acquisition of vehicle mobile measurement and tilt photogrammetry has been carried out. The theoretical accuracy of each data is analyzed and multisource data fusion is carried out, and then the real 3D model is reconstructed based on the fusion results. By comparing the effect of real scene 3D model based on tilt photography alone, the accuracy of multisource data fusion and the effect of model improvement are analyzed. In addition, automatic extraction technology of feature vector elements has been innovatively adopted based on vehicle movement measurement results, tilt aerial photography difference decomposition has been carried out combining with the data of domestic Beidou base station. It will improve wide application of real scene 3D modeling technology.