Accuracy of waste stockpile volume calculations based on UAV Photogrammetry

R. Künnapuu¹*, K. Kokamägi² and N. Liba³

¹Ministry of the Environment, Environmental Management Department, Paldiski rd 96, EE13522 Tallinn, Estonia
²Estonian University of Life Sciences, Institute of Agricultural and Environmental Sciences, Chair of Environmental Protection and Landscape Management, Kreutzwaldi 5, EE51006 Tartu, Estonia
³Estonian University of Life Sciences, Institute of Forestry and Rural Engineering, Chair of Forest and Land Management and Forest Industry, Kreutzwaldi 5, EE51006 Tartu, Estonia
*Correspondence: raunokunnapuu@gmail.com

Abstract:

In environmental supervision, it is necessary to measure waste piles volume to determine whether the activities of the waste manager comply with the established requirements. The aim of this research is to determine whether the model, formed from images collected with low-priced unmanned aerial vehicle (UAV) – not with Real Time Kinematic Global Navigation Satellite System (RTK GNSS) capability – is sufficiently accurate to carry out waste-related surveying. Data collection took place in spring 2021 at the Aardlapalu transhipment station in Tartu County. The objects of the research were an unscreened composting pile and a covered composting pile. In the fieldwork, terrestrial laser scanning and photogrammetric flight were carried out. The reference value was the volume of the model formed from the data of laser scan. The volumes of all models formed by the photogrammetric method were within the permissible difference of 10% provided by law. The most accurate results were obtained from the covered composting pile with an overlap of 70% × 70% and 21 ground control points (GCPs). Using these parameters, the absolute error of the model was 1.48 m³ and the relative error was 0.65%. The most inaccurate results were obtained from the unscreened composting pile with an overlap of 80% × 80% and 21 GCP-s. The research confirmed the hypothesis that sufficient accuracy to calculate waste piles volumes can also be achieved by using a cheaper UAV and camera and with software not specially designed for photogrammetry, design, and drawing.

Key words:

Analysis of flight parameters and georeferencing of images with different control points obtained by RPA

L.M.D. Santos¹*, G.A.S. Ferraz¹, M.T. Andrade¹, L.S. Santana¹, B.D.S. Barbosa¹, D.T. Maciel² and Giuseppe Rossi³

¹Federal University of Lavras, Department of Agricultural Engineering, University Campus, BR37.200-000 Lavras-MG, Brazil
²Federal University of Lavras, Department of Engineering, University Campus, BR37.200-000 Lavras, Brazil
³University of Florence, Department of Agriculture, Food, Environment and Forestry (DAGRI), Via San Bonaventura, 13, IT50145 Florence, Italy
*Correspondence: luanna_mendess@yahoo.com.br

Abstract:

New techniques for analysing the earth’s surface have been explored, such as the use of remotely piloted aircraft (RPA) to obtain aerial images. However, one of the obstacles of photogrammetry is the reliability of the scenes, because in some cases, considerable geometric errors are generated, thus necessitating adjustments. Some parameters used in these adjustments are image overlaps and control points, which generate uncertainties about the amount and arrangement of these points in an area. The aim of this study was to test the potential of a commercial RPA for monitoring and its applicability in the management of and decision-making about coffee crops with two different overlaps and to evaluate geometric errors by applying four grids of georeferenced points. The study area is located in an experimental Arabica coffee plantation measuring 0.65 ha. To capture the images, the flight altitude was standardized to a 30 m altitude from the ground, and a constant travel speed of 3 m s^-1 was used. The treatments studied were two combinations of image overlap, namely, 80/80% and 70/60%. Six points were tracked through Global Navigation Satellite System (GNSS) receivers and identified with signs, followed by an RPA flight for image collection. The obtained results indicated distinct residual error rates pointing to larger errors along Cartesian axis Y, demonstrating that the point distribution directly affects the residual errors. The use of control points is necessary for image adjustments, but to optimize their application, it is necessary to consider the shape of the area to be studied and to distribute the points in a non-biased way relative to the coordinate axes. It is concluded that the lower overlap can be recommended for use in the flight plan due to the high resolution of the orthomosaic and the shorter processing time.

Key words:

geometric error, orthorectification of images, overlap, photogrammetry, UAS (Unmanned Aircraft System)