Validation of Ortomosaics and Digital Terrain Models Using Photographs Obtained From a Not-Metric Digital Camera Attached to an UAV

This study aimed to verify the precision and accuracy of orthomosaic and Digital Terrain Model (DTM) automaticallygenerated by aerial photography program using aerial photographs taken with an Unmanned Aerial Vehicle (UAV)in the city of Goiás - GO, Brazil. The fl ight plan was drawn up in the progra...

ver descrição completa

Detalhes bibliográficos
Autores: Alves Júnior, Leomar Rufino, Côrtes, João Batista Ramos, Silva, Janete Rêgo, Ferreira, Manuel Eduardo
Formato: artículo
Estado:Versión publicada
Fecha de publicación:2019
País:Brasil
Recursos:Universidade Federal de Uberlândia (UFU)
Repositorio:Revista brasileira de cartografia - RBC (Online)
Idioma:portugués
OAI Identifier:oai:ojs.www.seer.ufu.br:article/49198
Acesso em linha:https://seer.ufu.br/index.php/revistabrasileiracartografia/article/view/49198
Access Level:acceso abierto
Palavra-chave:padrão de exatidão cartográfica
coordenadas tridimensionais
alvos pré-sinalizados
drones
standard cartographic accuracy
three-dimensional coordinates
pre-flagged targets
Descrição
Resumo:This study aimed to verify the precision and accuracy of orthomosaic and Digital Terrain Model (DTM) automaticallygenerated by aerial photography program using aerial photographs taken with an Unmanned Aerial Vehicle (UAV)in the city of Goiás - GO, Brazil. The fl ight plan was drawn up in the program E-mo-tion. The camera used was aCanon IXUS 220 HS, with spatial resolution of 12 megapixel, CMOS sensor equipped with type 1/2.3 “ (4000 x 3000pixels), 1.54 µm sensor element, and calibrated focal length of 4.382 mm, coupled to the UAV Swinglet CAM model. The fl ight was performed at approximately 716 meters above the ground with side and longitudinal overlappingof 40% and 70%, respectively, with a Ground Sampling Distance (GSD) of 25 cm. The photogrammetric block had9 tracks, with 62 photos, covering 889.39 ha. From those photos we generated two Orthomosaics and two DigitalSurface Model (DSM) with Pix4Dmapper software. The fi rst orthomosaic and DSM were generated using direct georeferencing without ground point supporting (i.e., without geodesic GPS survey). The second orthomosaic and DSMwere generated using, in addition to direct georeferencing, eight pre-signaled GPS points on the ground. The bias wasanalyzed by Student’s test (t) and accuracy by the probability of the chi-square (), following the standards set by theFederal Decree 89.817/84. The fi rst orthomosaic, without ground point supporting, was classifi ed as Class A on thescale 1/2.500, while the second orthomosaic, generated with ground point supporting, was classifi ed as Class A on thescale of 1/1.125. Thus, the fi rst orthomosaic presented a planialtimetric trend. Although the program has generated theDSM, DTM was evaluated due all the control points implanted on the ground. The DTM evaluated with and withoutground point supporting presented vertical displacement trend. It was evident the needing of ground point supportingfor making orthomosaics and DTM with UAV.