Open AccessBinarization of ESPI fringe patterns based on local entropy
Author(s) -
Mingming Chen,
Chen Tang,
Min Xu,
Zhenkun Lei
Publication year - 2019
Publication title -
optics express
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.394
H-Index - 271
ISSN - 1094-4087
DOI - 10.1364/oe.27.032378
Subject(s) - computer science , artificial intelligence , speckle pattern , preprocessor , pixel , speckle noise , entropy (arrow of time) , computer vision , pattern recognition (psychology) , cluster analysis , optics , physics , quantum mechanics
The fringe skeleton method is the most straightforward analysis method for phase extraction and widely used in dynamic measurement. Binarization is often required in this method. In the traditional binarization methods, filtering is often a necessary step prior to binarization due to the influence of intrinsic speckle noises in ESPI fringe patterns. In this paper, we propose a binarization method based on local entropy and fuzzy c-means (FCM) clustering algorithm. In this method, the pixels in the given ESPI fringe pattern are clustered into white fringes and black fringes according to their local entropy instead of the original intensity information. There is no need to perform the filtering preprocessing, because the intrinsic speckle noises are utilized as essentials. We evaluate the performance of our method by applying it to the computer-simulated and real fringe patterns. Experimental results demonstrate that the proposed method can achieve the desired binarization results, and the binarization results can give desired skeleton results.