Structure from motion technique of proximal sensing airborne data for 3D recontruction of extraction sites

Abstract

Nowadays digital photogrammetry based on Structure from Motion (SfM) algorithms represents a low-cost and rapid technique for the tridimensional detection and mapping of large areas of particular geologic-geomorphologic and environmental interest. In this case study, we use images captured from sensors mounted on a small and inexpensive aircraft for the 3D reconstruction of extraction sites. The aircraft, called RadGyro, is able to deal with morphologically complex zones, as the test area, located in the marble district of the Apuan Alps, the biggest and most exploited European extractive basin. The RadGyro is equipped with two digital cameras (full frame sensor and 35 mm fixed lens). Five GPS antennas and an Inertial Navigation System (INS) allow associating to each frame the camera position and its angular orientation. Ground Control Points (GCPs) acquired through traditional topographic surveys allow to improve the image exterior orientation as well as the SfM techniques allow detecting millions of tie points for an accurate alignment of photos. We finalize the photogrammetric processing by creating 3D point clouds, DSM, DTM and orthophotos. The results demonstrate how SfM algorithms, implemented in commercial software, emphasize the potentials of the RadGyro and are very useful in the photogrammetric processing.