Search

Due to color object surface interference, the distortion of captured sinusoidal fringe patterns in a colorful object measurement using fringe projection profilometry causes significant phase error and therefore shape measurement error. To address the problem, an intensity correction method is proposed to correct the acquired nonsinusoidal fringe of color object and aim to improve three-dimensional (3D) reconstruction accuracy. In this paper, first, the look-up tables (LUTs) and multi-fusion matrices (MFM) are established; secondly, finding the optimal indexing value to correct the color fringe pattern for improving recognition rates and sine; finally, the corrected pattern is separated into three phase-shifting fringe patterns, which can obtain demodulation phase by the three-step phase-shifting algorithm and then recover unwrapping phase. The simulation and experimental results demonstrate that the proposed method effectively reduces the phase error of color objects and achieves complete measurement of different color regions. [ABSTRACT FROM AUTHOR]