Avaliação da acurácia posicional de dados espaciais gerados a partir do uso da RPA

Abstract

Remotely Piloted Aircraft (RPA) is an example of equipment available in the market to meet the consumers’ need to obtain high-resolution aerial imagery. Considering the rapid acceptance of this equipment in the civilian environment and the advantages of the use of this technology for obtaining registration data, this study aims to evaluate the positional accuracy of spatial data obtained using a PHANTOM-3 RPA for purposes of updating registration databases. The study areas is located in the Glória campus of the Federal University of Uberlândia. The methodology included: flight planning; the collection of 52 control points, collected through post-processing static positioning using GNSS receptors and geometric levelling using a Ruide DL-202 Digital Level; creation of target flags and photogrammetric processing using the software Pix4Dmapper Pro - Trial. The accuracy evaluation was performed using orthomosaic and DTM obtained: without the use of control points (PC); using distributed PC; and, using PC grouped to the scale of 1:1.000 through tendency analysis, using inferential statistics, and precision analysis, through the Chi-Square test and the decree-law 89.817/ET-CQDG. In order to compare the discrepancies obtained between the test positional data and the reference data in different distributions of the PC, it was used the Analysis of Variance (ANOVA) and the Tukey test. The results obtained regarding the planimetric positional accuracy were approved for the images Distributed PC and Grouped PC. In addition, the Tukey test pointed out that these images do not differ statistically from each other. The fact that the images are from the same grouping might have favoured to obtain similar positional accuracy results. Therefore, RPA might be used for urban mapping, once it is capable of providing products with acceptable cartographic standards for cadastral surveys.