Digital Outcrop Model Generation from Hybrid UAV and Panoramic Imaging Systems

Abstract

The study of outcrops in geosciences is being significantly improved by the enhancement of technologies that aims to build digital outcrop models (DOMs). Usually, the virtual environment is built by a collection of partially overlapped photographs taken from diverse perspectives, frequently using unmanned aerial vehicles (UAV). However, in situations including very steep features or even sub-vertical patterns, incomplete coverage of objects is expected. This work proposes an integration framework that uses terrestrial spherical panoramic images (SPI), acquired by omnidirectional fusion camera, and a UAV survey to overcome gaps left by traditional mapping in complex natural structures, such as outcrops. The omnidirectional fusion camera produces wider field of view images from different perspectives, which are able to considerably improve the representation of the DOM, mainly where the UAV has geometric view restrictions. We designed controlled experiments to guarantee the equivalent performance of SPI compared with UAV. The adaptive integration is accomplished through an optimized selective strategy based on an octree framework. The quality of the 3D model generated using this approach was assessed by quantitative and qualitative indicators. The results show the potential of generating a more reliable 3D model using SPI allied with UAV image data while reducing field survey time and complexity.

Keywords

• point cloud
• 3D reconstruction
• spherical panoramas
• omnidirectional SPI cameras
• unmanned aerial vehicle