Modelo para evaluación de la exactitud de escáner láser terrestre - TLS

Abstract

SUMMARY The present work determines the accuracy of the points cloud captured by a Terrestrial Laser Scanner (TLS), with the aim of developing a statistical model, which allows the evaluation of the accuracy of the terrestrial scanner. This model is based on experimentation, with variation of reference parameters in direct measurements. Data is obtained through procedures that include capturing information at different types of distances, angles and surfaces (target). To give force to the validity of the model obtained, the data obtained are compared with respect to others acquired with Topographic Total Station. The mathematical and statistical development of the model uses the theory of the design of the experiment, where the measurements made in each of the scans are independent of each other, as is each target type. Proposing the implementation of the General Linear Model -GLM analysis, to adjust a mathematical expression to the average error (Euclidean distances between the observed and theoretical coordinates). The developed model allowed to determine the accuracy of the data acquired by the terrestrial laser, establishing that, for each additional meter in the distance of the scanner in relation to a study surface, the expected average error will increase between 0.033% and 1.5%.