Application of unmanned aerial vehicle multi-vision image 3D modeling in geological disasters
Shihan Bai, Shi-Han Bai, Shao-Yun Bai, Shao-Yun Bai · Journal of Information and Optimization Sciences · 2017
Frequent geological disasters in Yunnan province, southwest China, threaten the personal safety of local residents and the operational safety of construction projects. Multivision images of these areas can be quickly acquired using unmanned aerial vehicles (UAVs). From such images, a three-dimensional (3D) geographic information model can be quickly obtained using data processing. These models are useful for analyzing geological disasters. The present paper focuses on methods and key technologies for producing 3D models from UAV multi-vision images. Results showed that feature–grayscale matching adapted to image deformation and assured the matching precision. Finally, 3D models were constructed from two groups of image data. Compared with conventional imaging results, such as DEM, DOM and DLG, the 3D model was more precise, intuitive and accurate. Thus, geological workers can more clearly evaluate the geological environment and effectively predict and control geological disasters.