Extraction of Interference Fringe Information Based on Self-adaptive Restoration and Clustering via Distance Histogram

We propose two methods for efficient extraction of interference fringe information. The first method can restore the fringes in narrow strip areas through a series of operation: reducing noise, connecting broken points, extending short lines and polynomial fitting. The restoration process is self-adaptive and requires little user-defined parameters. The second method first employs distance histograms to cluster points belonging to the same fringes and then perform polynomial fitting, so that more accurate fringe information in large image areas such as curve shapes and extremum can be obtained. Experimental results show the reliability and accuracy of the proposed methods.