Model-Tracking-Based 3D Vibration Distribution Measurement of a Hovering Multicopter Using Stereo HFR Cameras

A comprehensive understanding of structural vibration distributions during multicopter flight, along with their correlation with flight trajectory and operational status, which are influenced by load conditions, battery levels, and weather variations, is crucial for ensuring flight safety. However, discrete sensors such as accelerometers and strain gauges face limitations in observing vibration distributions in multicopters due to restricted attachment points and their potential impact on flight performance. To address these challenges, we propose a model-tracking-based 3D vibration measurement method for flying multicopters using ground-installed stereo high-frame-rate (HFR) cameras, avoiding the need for onboard sensor installation. Our method utilizes digital image correlation (DIC) on stereo HFR images to achieve precise 3D vibration displacement measurements in the audio frequency range, leveraging thousands of tracked measurement points on a 3D multicopter model for whole-body vibration analysis. To validate our model-tracking method, we analyze stereo $1920\times 1080$ images at 500 fps in the frequency domain, capturing a hovering multicopter with four rotors rotating at hundreds of rps. Along with estimating its 6-DoF trajectory and rotor rotation speeds as operational status indicators, the 3D vibration distribution at 3518 tracked points was visualized by referencing a pre-measured 3D model of the multicopter, including 3D frequency component images at peak frequencies ranging from 0 to 250 Hz, which correspond to structural properties, control errors, and rotor-induced forced vibrations.