Intensity-ratio error compensation for triangular-pattern phase-shifting profilometry

Abstract

We present an intensity-ratio error-compensation method to decrease the measurement error caused by projector gamma nonlinearity and image defocus in triangular-pattern phase-shifting profilometry. The intensity-ratio measurement error is first determined by simulating the measurement with the triangular-pattern phase-shifting method with ideal and real captured triangular-pattern images based on the ideal and real gamma nonlinearity functions. A lookup table that stores the intensity-ratio measurement error corresponding to the measured intensity ratio is constructed and used for intensity-ratio error compensation. Experiments demonstrated that the intensity-ratio error compensation method significantly reduced the measurement error in the triangular-pattern phase-shifting method by 28.5%.

Binary coded triangular fringes for 3-D surface-shape measurement

“…Since the proposed technique is based on the projection of binary patterns instead of gray levels intensities, it is possible to expand the operating tolerance to ambient light and avoid camera sensor saturation. We also show that is not necessary the use of phase unwrapping, as proposed in [6][7][8][9][10][11][12] for recovering the unwrapped intensity-ratio map. Finally, we have demonstrated that the periodic measurement errors due to gamma nonlinearity and defocus of the projector and camera are eliminated with the proposed method.…”

“…An alternative procedure to the projection of sinusoidal fringes consists in projecting two consecutive patterns on the test surface; a constant illumination pattern and a linear gray-level pattern (along the x axis) or equivalently, a triangular pattern onto the object [5][6][7][8][9][10][11][12]. Then, the 3-D profile is calculated as a simple intensity ratio between the two patterns.…”

“…Although this technique has a reduced computational cost, the standard approach of triangular fringes projection techniques for 3-D surface-shape measurement has several drawbacks. First, due to the nonlinear response of the projector/camera system, the triangular patterns are not exactly triangular, and thus in practice, a tedious calibration (valid only for the specific combination of projector and camera used in the experiments) has to be performed to overcome the problem [6][7][8]. Second, if the intensity of the light incident on the object during measurement is greater than that during calibration, camera sensor saturation may occur depending on the gray levels in the computer-generated fringe patterns.…”

Super-Resolution Phase Retrieval Network for Single-Pattern Structured Light 3D Imaging

“…For phaseshifting multi-pattern SL methods, intensity variation information is coded in multiple patterns with phase shifts. Such intensity variations can be in the form of sinusoidal [7], [8], [9], [10], [11], [12], binary [13], [14], triangular [15], [16], hybrid [17], [18], or other shapes. Phase-shifting SL is the most representative method of SL because of its robustness and efficiency, and the sinusoidal pattern has the desirable properties of continuity and periodicity, and these properties make the sinusoidal pattern the most popular pattern for SL 3D imaging.…”

Investigation of accurate fringe analysis techniques for fringe projection profilometry

“…The phase was then, optimized by an iterative algorithm. To compensate phase error for triangular phase shifting profilometry, an intensity ratio error compensation method [182] was proposed to reduce the measurement error belonging to gamma non-linearity and image defocus. It uses a minimum of two triangular patterns to measure the 3D surface and hence, a simple intensity ratio method was used.…”