PSI - Issue 5

Francisco Barros et al. / Procedia Structural Integrity 5 (2017) 1260–1266 Francisco Barros et al./ Structural Integrity Procedia 00 (2017) 000 – 000

1265

6

Points whose displacement was outside the range of expected values were disregarded, as well as points which showed a large vertical difference to their rectified stereo matches, which should be indicative of a wrong match. Afterwards, a 2D interpolation was performed in order to form a continuous displacement field. The values for the displacement in the vertical direction were compared with results obtained from VIC-3D using the same subset and step sizes and the same region of interest (Fig. 4).

(a)

(b)

Fig. 4. Displacement in the vertical direction obtained through (a) the frequency domain based method (b) VIC-3D. Displacements are considered positive in the upward direction.

It can be seen that the displacement fields obtained through both methods are similar in magnitude and in spatial distribution. The range of displacement values differs slightly, likely due to points that were not matched correctly, and to the fact that the subsets may not have been exactly equal in both cases. The field obtained from the frequency domain method is less monotonic and appears to contain points whose calculated displacement does not seem to be correct, as it is quite different from the surrounding values. This is likely caused by erroneous correlation peaks, which will cause the appearance of wrong values in the shape reconstruction.