A new low-cost displacements monitoring system based on Kinect sensor

Abstract

We present a new system for structural health monitoring. In particular, the reason we have developed the system is related to structural monitoring during (potentially) destructive tests. Structural response is usually sensed by sensors such as accelerometers, LVDTs and all the instruments which need to be placed in contact with the structure in order to work properly. This can cause problems during (potentially) destructive tests, like the ones performed on shaking tables, an important class of experiments aimed at monitoring the behavior of certain structures under the action of a seismic force. Thus, in the present paper, we focus our attention on designing a new displacements monitoring system, which does not need to be in direct contact with the structure in order to work. The goal is to use a really promising technology in the field of 3D space monitoring to build a system suitable to replace classical sensors such as LVDTs and accelerometers, usually adopted in displacements monitoring. This technology is based on computer vision and low-cost sensors; this allows us to protect monitoring systems from damage due to contact with the structure under testing. Results show that our system is able to estimate the position of the selected points on a test structure with good accuracy, with respect to a traditional high-precision sensor.