Measurement of transparence topography based on principal component analysis

An algorithm based on the principal component analysis (PCA) is proposed to measure transparent surface topography. First, many frames of random phase shifted fringe patterns are captured. Second, the first and second principal components are extracted from the phase shifted fringe patterns by PCA. Then, the phase is calculated from principal components. Finally, the correct global phase sign is determined by the clustering method considering local fringe phase monotonicity. The experimental results show that the proposed method can directly extract the three-dimensional (3D) shape, which does not need to determine the amount of phase shift, without precise phase shifter. It reduces the experimental hardware requirements and improves its adaptability in profilometry.