Advanced processing and enrichment techniques for camera based dynamic testing and virtual sensing

Recent advances in high speed digital cameras, digital image processing and computational power are offering the possibility to involve camera based measurements to infer dynamic properties of mecha(tro)nic systems (eigenmodes, eigenfrequencies, local deformations, dynamical behavior of joints…). The advantages of camera measurements include a.o. contactless sensing (no added mass/damping introduced by sensors and cables), full field information (high spatial resolution), geometry information, and data in form of displacements. On the other hand, camera measurements bring in challenges in setting up a measurement campaign (camera settings and scene calibration, including lighting conditions), they produce a huge amount of data, and have a low temporal resolution in comparison with other (localized) sensors such as accelerometers, strain gauges, LDV. In this context, this PhD track focuses on reducing the impact of the limitations listed above by developing novel tools that allow to bring the added value of camera measurements to the field of dynamic testing. This includes novel techniques for feature detection and feature tracking (at subpixel resolution) including applied markers and the natural surface texture of the object under examination, automatic tools to determine the best settings for the camera and the scene (spatial resolution, sampling frequency, bit-depth, light, image stitching, sampling schemes if multiple cameras are used), virtual sensing to infer unmeasurable quantities such as operational loads, model updating, sensor fusion and other techniques aiming at higher acquisition rates without the need for expensive hardware (retrieve 3D displacements from one camera by introducing models, adding info coming from other sensors such as accelerometers, strain gauges, microphones). The developed techniques will be validated and demonstrated on industrially relevant cases, targeting a.o. design verification efficiency and end-of-line testing.