Noise-tolerant fringe evaluation using statistical fringe processing
In optical studies, fringe patterns often appear due to interference effects and the phase unwrapping process. The analysis of these fringe patterns is negatively affected by the presence of speckle noise, which tends to degrade the edge structures on the fringe pattern. In recent works, a statistical fringe processing algorithm that analyzes local spatial frequencies has been used in analyzing optical vortices and material sub-surface defects. In this study, the statistical algorithm is applied to noise-tolerant evaluation of fringes by determining the transverse position and boundary of the fringe pattern. The effects of a key parameter in the algorithm, the scanning window size, is demonstrated. The performance of the technique for various degrees of signal noise is also investigated. This technique may have applications in segmentation of fringe patterns.