High-speed three-dimensional measurement technique for object surface with a large range of reflectivity variations.

Abstract

It is a challenge for any optical method to measure object surfaces with a large range of reflectivity variations. Dark or saturated regions on fringe images may lead to phase and measurement errors. At present, the main defect of the existing methods is the low measurement efficiency. This paper presents a high-speed three-dimensional (3D) measurement for object surfaces with a large range of reflectivity variations. First, we set the intensity of the projected fringes to the maximum, which ensures a better signal-to-noise ratio (SNR) of fringe images compared to the traditional methods. Second, in order to achieve high-speed projection, we transform 12 fringes with 256 gray-value cosine variations to binary fringes and defocus the digital light procession to filter out the higher harmonics. Third, since manually adjusting the camera lens or exposure time is much more time-consuming and not conducive to improving the SNR, we use the color light projection of the projector itself to get multiple groups of fringe sequences with different brightnesses. Finally, the final group of fringe images, used for 3D reconstruction, are formed by choosing the brightest but unsaturated corresponding pixels from multiple groups of fringe images. Experiments verified that the proposed method has the advantages of high-speed measurement, low cost in hardware, high measurement accuracy, a simple algorithm, and ease of extending to several applications.