条纹规则化与希尔伯特变换相结合的三维面形测量方法  

Moving least square and Hilbert transform methods applied to Three-dimensional measurement

作　　者：边心田[1] 张勇兵 左芬[1] 程菊[1] 雷枫[1] BIAN Xintian;ZHANG Yongbing;ZUO Fen;CHENG Ju;LEI Feng(Jiangsu Key Construction Laboratory of Modem Measurement Technology and Intelligent System,Huaiyin Normal University,Huaian Jiangsu 223300,China;Graduate School of Syst. and Info. Eng., Univ. of Tsukuba, Ibaraki 305-8573 , Japan)

机构地区：[1]淮阴师范学院江苏省现代检测技术与智能系统重点建设实验室,江苏淮安223300 [2]筑波大学日本

出　　处：《激光杂志》2019年第4期41-44,共4页Laser Journal

基　　金：江苏省高等学校自然科学研究重大项目(No.17KJA460004);淮安市科技计划项目(HAC201701)

摘　　要：将移动最小二乘法和希尔伯特变换引入基于结构光投影的物体三维面形测量。移动最小二乘法拟合成的曲面具有精度高,光滑性好的优点,在测量过程中,利用移动最小二乘法得到离散条纹图的连续逼近函数,再结合希尔伯特变换具有90°相移的特性解出物体的相位信息,从而获得物体的三维面形分布。仅需要一帧变形条纹就可以重建被测物体的面形。与传统的傅里叶变换轮廓术相比,提出方法无需滤波操作,在局部阴影区域也可以解调出被测物体的相位,提高了测量的精度。计算机仿真和实验证明了提出方法的有效性。The moving least squares method ( MLSM) and the Hilbert transform method have been introduced into the three-dimensional shape measurement based on structured light projection. The fitting curve surface based MLSM has high precision and good smoothness. In measurement, MLSM is used to get the continuous approximate function of the discrete fringe pattern. Then According to Hilbert transform having the nature of 90° phase shift, The phase distri? bution can be drawn. As a result, the proposed method based on fringe projection one single deformed fringe pattern to reconstruct the tested object. Comparing with the Fourier transform profilometry ( FTP), this algorithm maintains a high spatial resolution for no filtering process. Furthermore, this method can demodulate the depth-related phase even in the local shadow of the fringe pattern. Both a computer simulation and the experimental result validate the feasibility of the proposed scheme.

关 键 词：三维测量 条纹分析 相位测量 希尔伯特变换 傅里叶变换轮廓术 移动最小二乘法 

分 类 号：TN29[电子电信—物理电子学]