Terrestrial laser scanning of civil engineering structures

Technika naziemnego skaningu laserowego stała się ostatnio bardzo popularną i dynamicznie rozwijającą się formą pomiarów geodezyjnych, w szczególności wykorzystywaną w dokumentacji obiektów architektonicznych oraz inżynieryjnych. Opracowane obiekty inżynieryjno-drogowe to most Świętokrzyski oraz wiadukt Tryon. Są to konstrukcje o dużym rozmiarze i rozpiętości, co implikuje zróżnicowane podejście do metody wykonywania pomiarów poszczególnych obiektów, przy uwzględnieniu wyboru lokalizacji i ilości stanowisk skanera oraz umiejscowienia tarcz celowniczych, a także wyboru metody rejestracji pozyskanych środowisk skanowania, właściwej dla danego typu obiektu oraz adekwatnej do otrzymanych wyników pomiarów.Recently, terrestrial laser scanning technology has become very popular technique of geodetic measurements. TLS has its application in road engineering geodesy. One of the advantages of 3D laser measurements are their high accuracy, resolution and number of points acquired in a very short period of time. One of the most important stages of the data post-processing is scan registration that means merging and fitting particular scans. This process can be done in a few ways, but also with different accuracy. There is a few methods of scan registration: using HDS targets, using some tie points from the clouds (so called cloud to cloud registration) and using both: targets and points (mixed method). Authors compared all this combinations on the example of two structures: Świętokrzyski bridge and Tryon viaduct. We have checked 3 ways of scan registration: by manual finding of points (with the RMSE in the range from 0.015 to 0.029 m with the mean RMSE 0.025 m), using HDS targets (0.003-0.006 m). Bridges are long and tall structures. In most cases there is also quite bid traffic. That causes, that very often, using HDS targets is impossible. There are also problems with the scanner range especially in case of long bridges. Another analysis was testing registration accuracy taking into consideration also the number of fitted points. We assessed, that minimal number of points necessary to registration with sufficient accuracy was about 100 to 200 thousands, what is about 1-2% of all points. In the paper we present the most important aspects of registration of data acquired with the use of terrestrial laser scanner