Real-time Pedestrian Dynamic-load Localization using Vision-based Motion Sensing

This paper presents a new method for analyzing the dynamic loading of a structure for structural health monitoring (SHM) using a computer-vision-based motion-sensing technique that obtains the pedestrian dynamic-load location. One of the main sources of error in the output-only SHM method comes from the assumption that the input (dynamic load) is white noise. In contrast, the input-output-based SHM method is rarely used because of the difficulty in measuring the dynamic load applied to the structure. Impact-load testing by employing an impact hammer is often used to control the dynamic load; however, the method requires additional pieces of equipment with traffic control. This paper presents a commercial off-the-shelf method to overcome such limitations using computer-vision techniques. The location of the pedestrian dynamic load is determined by applying motion-sensing algorithms. The proposed method does not require any additional equipment other than a commercial camera, which is costand labor-effective compared with conventional methods.