Concrete beams are used to construct bridges and other structures. Years of traffic overloading and insufficient maintenance have left civil infrastructure such as bridges in a poor state of repair. Therefore, the structures have to be strengthened. Many options for the reinforcement exist such as fibre-reinforced polymer composites and steel plates can be added. The efficacy of such methods can be effectively evaluated through fatigue load testing in which cyclic loads are applied to an individual structural member under laboratory conditions. During the fatigue test, the deflection of concrete beam is very important parameter to evaluate the concrete beam. This testing requires the measurement of deflection in response to the applied loads. Many imaging techniques such as digital cameras, laser scanners and range cameras have been proven to be accurate and cost-effective methods for large-area measurement of deflection under static loading conditions. However, in order to obtain useful information about behaviour of the beams or monitoring real-time bridge deflection, the ability to measure deflection under dynamic loading conditions is also necessary. This thesis presents a relatively low-cost and high accuracy imaging technique to measure the deflection of concrete beams in response to dynamic loading with different range cameras such as time-of-flight range cameras and light coded range cameras.
Due to the time-of-flight measurement principle, even though target movement could lead to motion artefacts that degrade range measurement accuracy, the appropriate sampling frequency can be used to compensate the motion artefacts. The results of simulated and real-data investigation into the motion artefacts show that the lower sampling frequency results in the more significant motion artefact. The results from the data analysis of the deflection measurement derived from time-of-fight range cameras have been indicated that periodic deflection can be recovered with half-millimetre accuracy at 1 Hz and 3 Hz target motion. A preliminary analysis for light coded range cameras is conducted on dynamic deflection measurement. The results demonstrate that the depth measurements of Kinect light coded range cameras are unstable, which implied that it is not sufficient to meet the accuracy required for the dynamic structural deflection measurement.