1,721,007 research outputs found
Static and dynamic features of Fourier transform profilometry: A review
Fringe projection techniques are very popular among the several profilometry methods proposed in literature. One of their main advantages is that most of them require one image of the object to retrieve its surface. From a theoretical point of view these techniques are based on a simple mathematical model which is at the base of the conversion of the measured quantity (usually the phase map of the image of the grid projected onto an object surface) into the height distribution of the object. However, the practical implementation of the system is not an easy task to be achieved. The mathematical models are often a simplification of the reality and it is valid only under some conditions. These discrepancies between the theory and the practice deal to a series of discussions about the quality of the measure of fringe projection techniques. This paper proposes a wide argumentation about the measurement uncertainty and the source of error which is influenced by. After a concise description of the triangulation model and of the basic idea of the system working, the measurement techniques are analyzed both from a static and a dynamic point of view, with a special attention to the uncertainty sources in the two measurement conditions and the possible improvements that may be applied to reduce their influence
Fourier-transform profilometry calibration based on an exhaustive geometric model of the system
In Fourier-transform profilometry (FTP), the height information is extracted from the phase distribution through triangulation; the relationship between the phase and height distribution depends on the system parameters such as the relative position of the projector and the camera, the fringe frequency and the reference plane position. In this paper, we propose a novel calibration approach for FTP that uses calibration planes to calculate the system parameters. The main innovation of this method is the application of an exhaustive geometric model of the FTP that expresses the phase-to-height relationship in the most general way with a camera and a projector not aligned; the aberration due to optics is also considered and compensated for. The obtained calibration data have a precise physic meaning and can be easily compared with the real system. Tests on both simulated and real data showed that the proposed method is robust, even in the case on non-negligible noise level
Comparison of eight unwrapping algorithms applied to Fourier-transform profilometry
Phase unwrapping is a task common to many applications like interferometry imaging, medical magnetic resonance imaging, solidstate physics, etc. Fourier transform profilometry (FTP) values the height distribution of object, elaborating the interference between a plane reference grating and a deformed object grating. Since the height information is extracted from the phase of a complex function, the phase unwrapping is a critical step of the process. Several unwrapping algorithms are proposed in literature, but applied to measurement technologies different from FTP. The purpose of this paper is to define the performances of eight different unwrapping algorithms applied to FTP optical scan method and to define the best one. The algorithms chosen are: Goldstein’s algorithm, quality guided path following method, Mask cut method, Flynn’s method, multi-grid method, weighted multi-grid method, preconditioned conjugate gradient method and minimum Lp-norm method. The methods were tested on real images acquired by a FTP scanner developed and calibrated for these experiments. The objects used vary from simple geometries, like planes and cylinders, to complex shapes of common use objects. Algorithms were qualified considering the phase unwrapping errors, execution time and accuracy of the shape of objects obtained from the scan method in comparison with real ones. The results show that quality guided algorithm best fits in FTP application
An experimental evaluation of the reliability of a damage localization algorithm based on FRF interpolation
Indexes for performance evaluation of cameras applied to dynamic measurements
Thanks to technology improvements, the applications of vision-based measurement to dynamic applications have been increasing in the last years. The available image resolutions and the high grabbing frequencies allow to acquire high-speed moving object with a good scaling factor and to perform dynamic analysis of vibrating items. Uncertainty analysis of vision-based measuring devices working in almost-static conditions was widely studied in literature, but the case of dynamic measurements still needs a further analysis. The measuring performances thus depend on the well-known parameters that affect the static performances (image resolution and contrast, processing algorithm, noise, etc.) but also on other factors, above all the exposure time and the camera-object relative motion, in terms of instantaneous velocity and acceleration. In this work, a performance analysis of imaging devices applied to dynamic measurements is proposed. The analysis aims to qualify the measurement uncertainty by some indexes, proposed in this work, and designed to quantify the motion effect on the acquired images and consequently the measurement uncertainty. These indexes are based on exposure time and Spatial Frequency Response (SFR) function, which is widely applied in literature and recommended in international standards for the image quality estimation in static acquiring conditions. Appropriate developments of SFR are proposed herein to obtain information on the image quality grabbed in dynamic conditions. The effectiveness of the proposed indexes are proved by several tests, where a target is moved with an harmonic law in controlled condition (varying its frequency and amplitude) and fixing different acquisition conditions in terms of lighting settings, diaphragm aperture, exposure time, etc
Experimental strain modal analysis for beam-like structure using distributed fiber optics
- …
