UAV RTK/PPK Method—An Optimal Solution for Mapping Inaccessible Forested Areas?

Martin Mokroš,Julián Tomaštík,Peter Surový,Ján Merganič,Alžbeta Grznárová

doi:10.3390/rs11060721

Martin Mokroš, Julián Tomaštík + Show 3 more

Open Access

https://doi.org/10.3390/rs11060721

Copy DOI

Abstract

Mapping hard-to-access and hazardous parts of forests by terrestrial surveying methods is a challenging task. Remote sensing techniques can provide an alternative solution to such cases. Unmanned aerial vehicles (UAVs) can provide on-demand data and higher flexibility in comparison to other remote sensing techniques. However, traditional georeferencing of imagery acquired by UAVs involves the use of ground control points (GCPs), thus negating the benefits of rapid and efficient mapping in remote areas. The aim of this study was to evaluate the accuracy of RTK/PPK (real-time kinematic, post-processed kinematic) solution used with a UAV to acquire camera positions through post-processed and corrected measurements by global navigation satellite systems (GNSS). To compare this solution with approaches involving GCPs, the accuracies of two GCP setup designs (4 GCPs and 9 GCPs) were evaluated. Additional factors, which can significantly influence accuracies were also introduced and evaluated: type of photogrammetric product (point cloud, orthoimages and DEM) vegetation leaf-off and leaf-on seasonal variation and flight patterns (evaluated individually and as a combination). The most accurate results for both horizontal (X and Y dimensions) and vertical (Z dimension) accuracies were acquired by the UAV RTK/PPK technology with RMSEs of 0.026 m, 0.035 m and 0.082 m, respectively. The PPK horizontal accuracy was significantly higher when compared to the 4GCP and 9GCP georeferencing approach (p < 0.05). The PPK vertical accuracy was significantly higher than 4 GCP approach accuracy, while PPK and 9 GCP approach vertical accuracies did not differ significantly (p = 0.96). Furthermore, the UAV RTK/PPK accuracy was not influenced by vegetation seasonal variation, whereas the GCP georeferencing approaches during the vegetation leaf-off season had lower accuracy. The use of the combined flight pattern resulted in higher horizontal accuracy; the influence on vertical accuracy was insignificant. Overall, the RTK/PPK technology in combination with UAVs is a feasible and appropriately accurate solution for various mapping tasks in forests.

Highlights

Forest inventory and mapping is essential for sustainability of forest ecosystems and forest management
We test the capability of the Unmanned aerial vehicles (UAVs) Real-Time Kinematic (RTK)/Post-Processed Kinematic (PPK) and demonstrate highly accurate spatial data comparable with data acquired using standard referencing approaches (GCPs)
The results of the RTK/PPK method were clearly more accurate and less prone to influence of these factors when compared to the ground control points (GCPs) approaches

Summary

Introduction

Forest inventory and mapping is essential for sustainability of forest ecosystems and forest management. Forests are often spread in less accessible areas, which might be considered hazardous for human access. Remote sensing techniques can be helpful in such cases and have been commonly used for this purpose [1,2,3,4]. Such studies mostly focus on estimation of canopy height since this is the most critical factor to predict risk [1,5,6,7,8,9] but they estimate other inventory parameters like tree basal area, volume (e.g., [3]) or ecosystem services [10]. Increasing availability of UAVs is contributing towards such studies [14]

Objectives

Methods

Results

Discussion

Conclusion