Coastal Mapping Using DJI Phantom 4 RTK in Post-Processing Kinematic Mode

Abstract

Topographic and geomorphological surveys of coastal areas usually require the aerial mapping of long and narrow sections of littoral. The georeferencing of photogrammetric models is generally based on the signalization and survey of Ground Control Points (GCPs), which are very time-consuming tasks. Direct georeferencing with high camera location accuracy due to on-board multi-frequency GNSS receivers can limit the need for GCPs. Recently, DJI has made available the Phantom 4 Real-Time Kinematic (RTK) (DJI-P4RTK), which combines the versatility and the ease of use of previous DJI Phantom models with the advantages of a multi-frequency on-board GNSS receiver. In this paper, we investigated the accuracy of both photogrammetric models and Digital Terrain Models (DTMs) generated in Agisoft Metashape from two different image datasets (nadiral and oblique) acquired by a DJI-P4RTK. Camera locations were computed with the Post-Processing Kinematic (PPK) of the Receiver Independent Exchange Format (RINEX) file recorded by the aircraft during flight missions. A Continuously Operating Reference Station (CORS) located at a 15 km distance from the site was used for this task. The results highlighted that the oblique dataset produced very similar results, with GCPs (3D RMSE = 0.025 m) and without (3D RMSE = 0.028 m), while the nadiral dataset was affected more by the position and number of the GCPs (3D RMSE from 0.034 to 0.075 m). The introduction of a few oblique images into the nadiral dataset without any GCP improved the vertical accuracy of the model (Up RMSE from 0.052 to 0.025 m) and can represent a solution to speed up the image acquisition of nadiral datasets for PPK with the DJI-P4RTK and no GCPs. Moreover, the results of this research are compared to those obtained in RTK mode for the same datasets. The novelty of this research is the combination of a multitude of aspects regarding the DJI Phantom 4 RTK aircraft and the subsequent data processing strategies for assessing the quality of photogrammetric models, DTMs, and cross-section profiles.

Optimization of UAV Flight Missions in Steep Terrain

“…On the other hand, the contribution of oblique images increases the consistency of the reconstructed surfaces. Tadia et al [33] improved the vertical accuracy of the photogrammetric model when they included oblique images within their nadiral dataset. Similarly, the authors of [34] state that the use of a tilted camera improves the robustness of the geometrical model.…”

Ground Control Point Distribution for Accurate Kilometre-Scale Topographic Mapping Using an RTK-GNSS Unmanned Aerial Vehicle and SfM Photogrammetry

“…The three investigations published in 2017 in Table 1 [37][38][39] each used customised data processing and analysis and produced maximum check point elevation errors of the order of 1 m. Weber and Lerch [40] did not report independent elevation errors. Forlani et al [32] and Taddia et al [41] each assessed different GCP configurations. Forlani et al [32] compared digital surface models (DSM) using Root Mean Squared Error (RMSE) statistics to determine the quality of SfM photogrammetry outputs based on different GCP scenarios.…”

“…Forlani et al [32] compared digital surface models (DSM) using Root Mean Squared Error (RMSE) statistics to determine the quality of SfM photogrammetry outputs based on different GCP scenarios. Taddia et al [41] surveyed 2 km of coastline with a DJI Phantom 4 RTK and compared z-axis RMSE in PPK results. From the investigations in Table 1, assessments of the DG method suggest that: (i) DSMs georeferenced with no GCPs produce consistently higher z-axis RMSE than those with one or more GCPs [32].…”

Boosting the Timeliness of UAV Large Scale Mapping. Direct Georeferencing Approaches: Operational Strategies and Best Practices

“…Among the main scientific papers focused on the use of this platform, Ref. [33] compared the performances of a Phantom 4 Pro and a Phantom 4 RTK; in [68,69], the authors tested different direct georeferencing strategies for the monitoring of a coastal area; Ref. [70] created a test field with points of known coordinates to evaluate the accuracy of different GNSS flight modes; and finally, Ref.…”

“…Two different set of parameters have been used: one calculated months before the analyzed dataset ( [65], Table 7, second column) and one obtained through an on-site self-calibration in the same area ( Table 7, third column). The pre-calibration provided by DJI (Table 7, first column) was not used due to the poor results obtained in previous experiences (as previously mentioned) and other considerations highlighted in [69].…”