Open AccessAdaptive noise reduction method for DSPI fringes based on bi-dimensional ensemble empirical mode decomposition
Author(s) -
Yi Zhou,
Hongguang Li
Publication year - 2011
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.19.018207
Subject(s) - speckle pattern , noise reduction , hilbert–huang transform , speckle noise , noise (video) , optics , reduction (mathematics) , mixing (physics) , computer science , mode (computer interface) , interferometry , materials science , artificial intelligence , physics , computer vision , mathematics , image (mathematics) , geometry , filter (signal processing) , quantum mechanics , operating system
Digital speckle pattern interferometry (DSPI) fringes contain low spatial information degraded with speckle noise and background intensity. The denoising technique proposed recently based on bi-dimensional empirical mode decomposition (BEMD) could implement noise reduction adaptively. However, the major drawback of BEMD, called mode mixing, has affected its practical application. With noise-assisted data analysis (NADA) method, bi-dimensional ensemble empirical mode decomposition (BEEMD) was proposed, which has solved the problem of mode mixing. The denoising approach based on BEEMD will be presented, compared with other classic denoising methods and evaluated both qualitatively and quantitatively using computer-simulated and experimental DSPI fringes.