ASSESSMENT OF THE ACCURACY AND PRECISION OF A NEAR-REAL-TIME PROCESSING METHOD OF LASER SCAN DATA APPLIED ON CONCRETE STRUCTURES

Abstract

By now, the use of terrestrial laser scanning for deformation measurements of large concrete structures is well established. To tackle the time consuming processing of the large amounts of data that are the result of the scanning process, a 'Quick and Dirty Method' (QDM) was developed and documented. In order to quickly derive a certain number of cross sections at predefined locations, this QDM implies no cleaning up of the point clouds and no registration, but only primitive modeling of parts of the structure. This allows a near-real-time processing of the data, with resulting deformation values that are similar to the values obtained by total station measurements. The principle idea was to assess the possibility of using laser scanning in time critical situations for which until now, only conventional total station measurements are used. The potential of the QDM was demonstrated with a specific test case: namely the monitoring project of a sewage purification plant. Furthermore, the degree of detail that can be reached by using laser scanning surpasses the possibilities of total stations. Moreover, the visualization of deformations by the QDM is much more comprehensive than the results that can be obtained by using a total station. However, the previous method was only tested in one particular case study of a sewage purification plant. In a real-time situation, it is not possible to obtain the number of measurements needed to produce statistically correct statements on the quality of the data. Consequently, a thorough assessment of the accuracy and precision was impossible. Besides, as the range errors depend on the measured distance, the incidence angle of the laser beam, the material characteristics of the object and the type of laser scanner, these parameters have to be taken into account to assess the conditions in which the QDM can be used and to quantify its performance. In the more advanced research at hand, a concrete test specimen was scanned from six different distances from the scanner. In order to assess the influence of the incidence angle on the point cloud, the specimen was also rotated over five different angles and then translated over about one centimeter (measured with a dial gauge). For each position of the test specimen, 120 scans were performed with a point spacing of 2 mm by 2 mm. This resulted in approximately 720 scans for each scanner to object distance. This paper elaborates on some of the most important results for the Leica ScanStation C10.