Determinación de precisión entre nubes de puntos de alta densidad, obtenidas con sensores LIDAR

Abstract

The present work makes a comparison and determines the precision of data captured by airborne sensors on UAVs with Lidar technology, high-density point clouds were collected with the DJI Zenmuse L1 sensor and YellowScan with the Explorer, Mapper and Ultra models, in the same zone. The data for comparison were taken in an area of approximately 5 hectares, including the buildings of the Roosevelt Institute and District University, specifically the property of the Faculty of Environment and Natural Resources. The development is divided into six phases, where digital models of the terrain are generated. Profiles are drawn to evaluate position data, particularly coordinates of representative points of the ring road and their horizontal signaling marks on the road section are established. To analyze the result of the precision obtained, a product of the high-density point clouds for each of the sensors, the ANOVA statistical model of one factor for elevations and two factors for coordinates was used. The results obtained were: with the variability coefficient, concluding that the point cloud with the least data dispersion, that is, the greatest precision with respect to the others, is that of the Explorer model sensor with a percentage of variability for X, Y coordinates of 8.6058% and 0.0883% in levels.