基于调制度分析的加权最小二乘位相展开方法

提出将条纹光强调制度分析和加权最小二乘法相结合进行二维位相全局展开的一种方法，在Ghiglia和Romero提出的基于离散余弦变换的加权最小二乘位相展开算法的基础上，根据条纹图像调制度分析确定加权最小二乘算法中的权矩阵，对调制度低的位相数据赋予低的权值，可以有效屏蔽无效位相数据对邻近区域的不利影响，该方法更充分利用条纹图像信息，算法稳健，可以实现完全自动的二维位展开，对实验数据的处理取得了较满意的结果。

Weighted Least Square Phase Unwrapping Utilizing Intensity Modulation Analysis

A new two dimensional phase unwrapping technique for fringe analysis is proposed and verified by experiments. The technique is based on the algorithms of weighted least square phase unwrapping using fast discrete cosine transforms and iterative methods, which were proposed by Ghiglia and Remero in 1994. Intensity modulation is calculated and used to create the weight matrix. Invalid areas of wrapped phase are weighted by small values. Thus their unfavorable effects on the final unwrapped phase map can be limited. The phase unwrapping process of fringe pattern can be achieved without any priori information or manual intervention. The iterative algorithms are proved to be highly robust.