Quality assessment of products generated by survey of area using remotely piloted aircraft and control points

Abstract

Space technologies have allowed productivity gains and an improvement in cost-benefit, being the case of most activities that depend on a spatio-temporal analysis of their occupation on the Earth's surface. The use of georeferenced high spatial resolution images together with Geographic Information Systems (GIS) contributes significantly to environmental studies. Even with all technological advances, georeferenced images still have positional uncertainties and, therefore, their quality must be evaluated in order to increase the accuracy of the products generated. In Brazil, this assessment can be made from the Cartographic Accuracy Standard for Digital Cartographic Products (PEC-PCD) defined by the Technical Specification for the Acquisition of Vector Geospatial Data (ET-ADGV) (DSG, 2011), being based on the Standard of Cartographic Accuracy (PEC) defined in Decree-Law no. 89,817 of 1984 (BRASIL, 1984), which regulates the classification of products as to their accuracy. The aim of this work was to use control points collected with Real-Time Kinematic positioning system to reduce positional uncertainties and errors of georeferenced images in order to contribute to the search for functional and dynamic forms of management, control and environmental recovery. The study area is located in the northeastern region of the municipality of Rondonópolis - MT, between coordinates 16 ° 6'57.91 "S and 54 ° 29'3.69" O; 16 ° 7'1.04 "S and 54 ° 29'42.50" O, with an area of approximately 77 hectares. For the mapping of the area, aerial photographs of high spatial resolution were used, obtained through an RPA model DJI Phanton Pro 4 and six control points were collected in the area using a Topcon Hiper V GNSS receiver. Data processing was performed using the Agisoft Metashape Professional software (Agisoft, 2018). As a final product, the Digital Elevation Model - DEM was obtained together with the level curves, and for comparative purposes, two processing reports were generated, one before the insertion of the support points and the other containing the points. It was concluded that the insertion of the support points significantly decreased the projection errors contained in the referenced images, and before the total error in the x, y and z axes was estimated at 8.507m and after the insertion the error became 0.191 m. From this product with refined accuracy, it was possible to identify the erosions of the study site efficiently, generating results and an interesting routine for the management of degraded areas, giving a way for research centers to use this method in the management of places that suffered from plant suppression and present erosion from this suppression.