A new algorithm for large surfaces profiling by fringe projection

Abstract

7 pagesA method for profiling surface objects based on the fringe projection method and a phase shifting algorithm is described. The application of this method to large surfaces is problematic since the calibration step requires the use of a reference plane as large as the object. A new algorithm based on least-squares method has also been developed to bypass this calibration step, and so the use of a reference plane. First experimental results on a carbon panel and on a parabolic aerial are presented to show the validity of the proposed algorithm. Accuracy of 1mm has been obtained for an object of 1m and 20 cm long, while sensitivity has been proved to be of the order of 100 m

Nonlinear calibration for generalized fringe projection profilometry under large measuring depth range

“…Yin et al [7] presented a robust calibration approach based on the stereo vision model of the FPP scheme and bundle adjustment strategy. A new calibration for large surface profile was presented by Pavageau et al [8] based on the least-squares method. In the calibration proposed by Liu et al [9], system distortions for each detection location were characterized individually, which improves the measuring accuracy effectively.…”

A study on carrier-removal techniques in fringe projection profilometry

“…The phase-to-height conversion is also referred to as the calibration of a system. Several researchers [6][7][8][9] have proposed direct calibration methods, in which the phase-to-height relationship is obtained from system geometrical parameters without the necessity to remove the carrier phase component. Zhou and Su [6] proposed implementing the phase-to-height conversion based on two geometrical parameters obtained by measurement of at least three different parallel reference planes.…”

“…However, estimation of the unknowns is computationally involved and the minimization algorithm relies on manual intervention to prevent it from being trapped in an erroneous solution. A more practical strategy was subsequently proposed by Pavageau et al [8]. Instead of using a minimization process to quantify all the system geometrical parameters, Pavageau et al suggested estimating only three of the most important parameters from an unwrapped phase map and directly measuring the rest.…”

Time resolved profile measurements for ultra-fast vibrating objects

“…Phases of the fringes recorded by the image acquisition system at a different viewpoint are analyzed to retrieve the 3D shape. With proper phase-extraction schemes [1][2][3][4][5] and calibrations, depth accuracy better than one part in ten thousandth of the field of view can be achieved [6][7][8][9][10][11].…”