Two-Dimensional DIC Using Moving Camera and Depth Sensor

The use of vision-based techniques for structural health monitoring (SHM) tasks is an interesting solution mainly due to the possibility to obtain field measures in a relatively cheap way. However, the optimal placing of cameras may be difficult in real scenario with hardly accessible structures and the use of vision systems mounted on robots like unmanned ground or aerial vehicles has become a possible solution to overcome this limitation. However, when digital image correlation (DIC) is used to measure displacements, the movement of robot gives rise to perspective distortions that corrupt the obtained displacement fields. In this article, a technique based on the combined use of a camera and a depth sensor for the compensation of such distortions is proposed for the case of planar surface measure. A system made of a camera and a depth sensor fixed on the same support is used to record the measurement target. The relative position between camera and depth sensor is obtained by calibration; therefore, information from depth sensors can be used to reconstruct the relative pose between camera and target surface, giving the possibility to compensate perspective distortions. The proposed system has the advantage to be more compact and less heavy than other solutions like stereo camera systems, being suitable for those applications in which the system must be moved. In this article numerical and experimental tests are presented for the problem of dynamic crack aperture evaluation. Tests showed good performances of the proposed technique, with errors in aperture measure less than 0.01 mm when the system is 1 m far from the target surface.