Accuracy of Photogrammetric Products Generated by Remotely Piloted Aircrafts in Rough Reliefs

Abstract

Despite the popularization of the use of remotely piloted aircraft (RPA) as tools for the production of photogrammetric and digital cartographic products, little is said about the accuracy of such products in the context of rugged terrain, where the abrupt difference in altitudes leads to greater complexity. relief modeling and consequently in the generation of orthophotos. The aim of this paper is to present the evaluation and classification based on the PEC-PCD (Cartographic Accuracy Standard for Digital Cartographic Products) of orthophotomosaics and digital surface models (DSM) generated for the same mining area. For this study we used RGB (non-metric) images with GSD (Ground Sample Distance) estimated at 2.45 cm, and 80% / 80% overlap, captured by multirotor RPA on two identical flights performed on different dates, for each 15 pre-signaled targets were used from which the X, Y and Z coordinates were collected with the aid of GNSS RTK equipment. Five experiments were performed, varying the number of GCP (Ground Control Points) and maintaining the number of CP (Check Points). The products (orthophotomosaic and DSM) generated with the different GCP configurations were evaluated based on the PEC-PCD and, after the analysis of the results obtained, it was possible to verify the scale variation in which the products fit, this fact was attributed to the quantity and GCP arrangement (geometry). In general, products generated with 6 and 8 GCP presented similar accuracy levels and were classified as Class A for the 1: 1000 scale.