*3.4. The TLS Technique*

The TLS technique permits receiving a quasi-constant, three-dimensional presentation of the observed objects. The precision of the measuring effects depends mainly on the sight length, i.e., the distances between the elements of the tested object and the scanner. The number of the TLS applications is growing rapidly [21,22,30]. TLS measurements are related to determining the distance based on a difference between the reflected and emitted laser beam. Data on the obtained distances and the reflected signals are transformed into coordinates in real time. The operation principles are described in detail by Rodriguez [31].

FARO Focus 3D (Lake Mary, FL, USA) and Trimble TX8 scanners (Sunnyvale, CA, USA) (Figure 8) were applied for the bridge deflection testing. These scanners measure almost 1 million points per second and allow building a cloud consisting of even a billion points. The manufacturer declares that the measurement precision of the FARO Focus 3D scanner is on the level of ±2 mm, while the angular resolution is equal to 0.16 mrad. The maximum scanning range is up to 130 m. The maximum scanning range of the Trimble TX8 scanner is 120 m, but in favorable conditions, it may even extend to 340 m. The laser wavelength is 1.5 μm, and the scanning frequency is 1 MHz. The declared measurement accuracy is ±2 mm, and the angular resolution is 0.07 mrad. A comparison of catalogue data shows that the Trimble TX8 scanner has twice the angular resolution as the FARO Focus 3D scanner. Anigacz et al. [32] applied the FARO laser scanner during the first attempt to check the applicability of the TLS technology to bridge deflection testing. However, the results indicate that the FARO laser scanner used did not allow for obtaining accurate and satisfactory data. The quality of the scanning results of the bridge deflections was close to the maximum range of the scanner (120 m). For this reason, the quality of scans (density of the points) was too weak to obtain the bridge deflections with satisfactory accuracy.

**Figure 8.** Side view of the bridge measurement sets for laser scanning and photogrammetry.

The Trimble TX8 scanner allows obtaining the point clouds for the unloaded and loaded state clouds with a 5.7 mm resolution at a 30 m distance. This gave about 500 million measurement points for a single scan. The unloaded state was registered on five measuring spheres to the loaded state with accuracy under 1 mm using Trimble RealWorks Advanced Plant software version 10.2 [33]. Visualization of the deviation cloud between the unloaded and loaded state was obtained through an authoritative optimization approach that was then used for comparative analysis in Trimble RealWorks Advanced Plant 10.2. The software permits for the autointerpolation of the center of the measuring shields and spheres, which were also used in the photogrammetry technique (Figure 7).

It is important that measuring scanning sessions were compatible with each other; therefore, succeeding scans should be matched to selected bridge points. This approach significantly reduces the measurement time on the structure.

When TLS is used, the compared point clouds are recorded in a common coordinate system, or both are georeferenced if a global reference system exists. In the analyzed case, one point cloud represents the bridge state without load and the other cloud represents the state with a static load. There are two cases of point identification, i.e., characteristic points of the structure and points signaled by markers recognized by the software (e.g., checker, spheres). It should be emphasized that much better accuracy in determining the location of points is obtained when they are signaled (as in the tested bridge).

In this case, the results of laser scanning were additionally presented in the form of color deviation maps. The generation of deviation maps consists of comparing a georeferenced point cloud with a 3D model. Based on differences in the distance between the 3D model and the point cloud, the system allows you to give colors to the cloud with a given RGB value depending on the reference distance (i.e., the surface of the 3D model). In the analyzed case, navy blue was used for 0–4 mm deformations, green-yellow was used for 24–32 mm deformations, and red was used for 40–50 mm deformations. The color-defined deviation maps can be indicated by the operator performing the study.
