Dynamic phase measurement by clustering method

Phase measurement is a key step in quantitative optical metrology. While phase shifting technique is widely applied for accurate and reliable static or semi-static phase measurement, Fourier and wavelet transforms are often employed for high speed dynamic phase measurement. In our previous papers, the authors had proposed an alternative clustering method for dynamic phase measurement. The proposed method utilizes the phase clustering effect and the prior knowledge of the speckle field to extract the deformed phase map from one single deformed speckle pattern. The clustering method, however, may fail at area with abundant noise and large phase gradient. In this paper, we improve the clustering method by incorporating an advanced phase filtering methods for wrapped phase filtering. The reconstructed wrapped phase map is with very good quality and ready for phase unwrapping with any simple unwrapping algorithms. The basic ideas and the implementation approach will be described in details. Several examples based on shearography and holographic interferometry will be presented. Comparisons between the proposed method and phase shifting method will be made. The results demonstrate the accuracy and robustness of the integrated dynamic phase extraction method. The integrated phase retrieval method proposed here has great potential to simplify optical setup for dynamic phase measurement.