Análisis de la precisión en levantamientos realizados por UAV respecto al campo de visión en cámaras no métricas de bajo costo y la densidad de puntos de control.

Abstract

Abstract: The purpose of this research is to determine the accuracy in lifting carried out by unmanned aerial vehicles (UAV) regarding two variables: the field of vision and the density of control points, using non-metric cameras. There were two overflights with drone; the first one was done with a field of medium mink and a height of 50 meters and the second, taking into account a normal field of vision and a height of 40 meters. Furthermore, two digital models were generated per each overflight through the Structure from Motion software so that the geometry and density of control points can be varied. Via this research, it was possible to establish that the average field of view generates smaller errors in coordinates and the normal field of vision smaller errors in pixels. According to the geometry it was determined that the trapezoidal form produces bigger errors than the triangular form. Regarding flight height, it was established that at 50 meters the coordinate errors were less than 40 meters and that the errors in pixels were higher.