Still-camera multiview Spectral Optical Flow Imaging for 3D operating-deflection-shape identification Domen Gorjup a , Janko Slaviˇ c a,* , Aleˇ s Babnik a , Miha Bolteˇ zar a a University of Ljubljana, Faculty of Mechanical Engineering, Aˇ skerˇ ceva 6, 1000 Ljubljana, Slovenia Cite as: Domen Gorjup, Janko Slaviˇ c, Aleˇ s Babnik, Miha Bolteˇ zar, Still-camera multiview Spectral Optical Flow Imaging for 3D operating-deflection-shape identification, Mechanical Systems and Signal Processing, Volume 152, 2021, 107456, ISSN 0888-3270, https://doi.org/10.1016/j.ymssp.2020.107456. Abstract To measure high-frequency 3D vibrations, multi-camera, high-speed imag- ing hardware is normally required. An alternative using still-frame cameras was recently introduced with the Spectral Optical Flow Imaging (SOFI) method. In this research, the SOFI method is extended to multiview measurements of spa- tial operating deflection shapes. This is achieved by utilizing harmonically con- trolled illumination to perform an analogue Fourier transform on image-intensity data in multiple camera views. The obtained multiview displacement spectra are combined with geometrical data to perform frequency-domain triangulation and reconstruct spatial deflection shapes. By introducing additional camera views into the image-based measurement, its field of view is extended and the signal- to-noise ratio of the final result is increased. For linear, time-invariant mechan- ical structures under stationary excitation, full-field 3D measurements of high- frequency vibrations can be performed using a single still-frame monochrome camera. The proposed method identifies displacements in the frequency domain directly on the camera sensor, resulting in orders-of-magnitude smaller data sizes and post-processing times compared with conventional multiview image-based methods. Keywords: Spectral Optical Flow Imaging, full-field displacement measurement, single camera, multiview, still camera, frequency domain triangulation * Corresponding author Email address: janko.slavic@fs.uni-lj.si (Janko Slaviˇ c) Preprint submitted to Mechanical Systems and Signal Processing December 2, 2020