PSI - Issue 37

Francisco Barros et al. / Procedia Structural Integrity 37 (2022) 880–887 Barros et al./ Structural Integrity Procedia 00 (2019) 000 – 000

882

3

The 3D system setup consisted of two Bosch DINION IP Ultra 8000 MP cameras with a 75 mm lens, acquiring images at full resolution (4000 x 3000 px), mounted on tripods near the windows of the third and fourth floors of the building, as can be seen in Fig. 1 and Fig. 2, such as to capture images of the bridge from the top. The cameras, along with the computer used to control the acquisition, were connected to a common Ethernet switch. The 2D setup consisted of a USB camera, an iDS uEye UI-3370CP model (2048 x 2048 px of resolution), with a 150 mm lens, mounted on a tripod near a ground floor window (Fig. 1) and capturing the bridge from the side, connected to a laptop computer.

Fig. 2 . Bosch DINION camera capturing the bridge’s surface from above.

The region of the bridge where the displacement was measured was chosen because it could be seen clearly from all locations where cameras were set up. This was decisive especially for the 2D test, since the line of sight from the ground floor to some of the extension of the bridge was obstructed. The bridge’s surface as seen from above has an irregular contrasting pattern on a scale compatible with the application of DIC, as confirmed by preliminary tests, therefore it was not necessary to prepare the surface with a custom speckle pattern for the 3D test. For the 2D test, it was observed in preliminary tests that some natural features of the bridge could be followed with DIC from the side, so fastening a target to the side of the bridge was not needed for the purpose. On the 3D system, the cameras were 4 m apart and around 30 m away from the region of interest, resulting in a stereo angle of approximately 3° and a field of view of about 3.5 m along the length of the bridge. The camera for the 2D system was placed 13 m away from the monitored section, capturing 1.4 m along the length of the bridge.

2.3. Image acquisition

Before the test, calibration images were captured for both the 3D and the 2D systems. On the 3D test, calibration images were acquired with a checkerboard pattern printed on an A0 size PVC panel with aluminium profiles fastened to its back in order to prevent it from bending. On the 2D test, the camera could not be oriented with its axis perpendicular to the measured surface, so a single calibration image of a smaller checkerboard pattern held against the surface was taken in order to apply a perspective correction later.