Optical Reconstruction of High-Speed Surface Dynamics in an Uncontrollable Environment

摘要: The ability to communicate with our voice can be regarded as the concatenation of two processes “phonation” and “modulation.” These take place in larynx palatal oral region, respectively. During phonation audible primary signal is created by mutual reaction vocal folds exhaled air stream lungs. underlying interactions masses, fluids acoustics have yet identified understood. One part signal's acoustical source are vortex induced vibrations, e.g., Coandaeffect stream. development these vorteces determined shape 3-D movements larynx. Current clinical vivo research methods for do not deliver data satisfactory quality fundamental research, an endoscope limited 2-D image information. Based hereupon, a few improved been presented, however delivering only selective information, either single point or line. This stands contrast motions entire fold surface. More complex imaging methods, such MRI, information real-time. Thus, it necessary develop easily applicable, more examination method, which allows surfaces obtained. We present method calibrate reconstruction setup including laser projection system high-speed camera. designed miniaturization application mind. generates divergent grid 196 dots diffraction gratings. It calibrated planar calibration target through homography. In general, reconstruct topology surface at high frame rates (up 4000 frames per second) uncontrollable environments, given lighting situation (little no ambient light) varying texture (e.g., grade reflection) human particular, this measures dynamics during phonation. Applied synthetic data, shown robust (error approximately 0.5 ) regarding noise systematic errors. Experimental gained linear -stage proved that reconstructs coordinates points error 15 . was applied exemplarily porcine artificial folds' Local differences asymmetry between left right dynamics, well global parameters, opening closing speed maximum displacements, were quantified.