W pracy przedstawiono wyniki badań dotyczących ujawniania, klasyfikacji i oceny błędów o charakterze jakościowym w pomiarach odległości dokonywanych skanerami laserowymi 2D stosowanymi w nawigacji robotów mobilnych. Na podstawie analizy fizycznych podstaw pomiaru określono przyczyny oraz warunki i zakres występowania tego rodzaju błędów dla dwóch typów skanerów laserowych o odmiennych właściwościach.
EN
This paper reports the results of research concerning recognition and classification of the qualitative-type measurement uncertainty in 2D laser scanners used on mobile robots for navigation. The main source of the qualitative-type uncertainty is the so-called ”mixed pixels” effect. This effect has been investigated experimentally for two different classes of the laser range sensors, and explained by analyzing the physical phenomena underlying their operation.
Using a lower-cost laser scanner for generating accuracy in 3D point-cloud has been a concern because of economic issues; therefore, this study aims to create a 3D point cloud of a target object using a low-cost 2D laser scanner, Hokuyo UTM 30LX. The experiment was carried out in November 2019 with 16 single scans from 8 different view points to capture the surface information of a structure object with many intricate details. The device was attached to a rail, and it could move with stable velocity thanks to an adjustable speed motor. The corresponding 16 point-clouds were generated by using the R language. Then, they were combined one by one to make a completed 3D point cloud in the united coordinate system. The resulted point cloud consisted of 1.4 million points with high accuracy (RMSE = ±1:5 cm) is suitable for visualizing and assessing the target object thanks to high dense point-cloud data. Both small details and characters on the object surface can be recognized directly from the point cloud. This result confirms the ability of generated the accuracy point cloud from the low-cost 2D laser scanner Hokuyo UTM 30LX for 3D visualizing or indirectly evaluating the current situation of the target object.
JavaScript jest wyłączony w Twojej przeglądarce internetowej. Włącz go, a następnie odśwież stronę, aby móc w pełni z niej korzystać.