一种复合编码的三维数字成像方法

Approach for Three-Dimensional Imaging Based on Hybrid Encoding

为了解决物体表面存在不连续或大梯度的三维数字化问题,提出了一种结合点阵编码和条纹编码两种方式的复合编码三维数字成像方法。该方法综合了点阵编码和条纹编码两种编码方式的优点,利用点阵编码三维数字成像绝对测量的特点,以点阵编码获取物体较低分辨率的三维深度像,并以此确定条纹级次,控制条纹解码的相位展开过程。建立了基于这种复合编码方法的三维数字成像理论模型,并通过实验验证了该方法可有效地克服相位模糊和误差传播问题,同时该方法的物体深度测试范围和测试对象范围均大于单一编码方式。

A novel three-dimensional （3D） imaging method based on hybrid encoding is proposed to solve the 3D profilometry of objects surface with large height discontinuities and/or with very steep slope. By taking advantage of both point-array encoding method and fringe encoding method a hybrid encoding technique is constructed in which the point-array encoding method is used firstly to get range image, decide the fringe order, and control the further phase unwrapping processing. A theoretical model of 3D reconstruction based on the hybrid encoding is established. The experimental results are presented to demonstrate that the problem of the phase ambiguity and the error propagation can be avoided with proposed method. With proposed method, the range of height of tested object surface and class of surfaces can be extended compared with single encoding method.