Advanced processing of optical fringe patterns by automated selective reconstruction and enhanced fast empirical mode decomposition

Hilbert-Huang processing and analysis of complex fringe patterns

“…Beside significantly shortening the computation time, more accurate and meaningful estimation of the bidimensional IMFs (BIMFs), representing image features at various spatial scales is obtained in many cases. For further computation time reduction the Enhanced Fast EMD (EFEMD) algorithm with order-statistic filtration implemented using the morphological operation of dilation was proposed in [59]. To reduce the computational load the EFEMD algorithm estimates the filter window width counting the number of extrema and assuming its quasi-homogenous distribution.…”

“…In other words the operator decided which value of the amplitude modulation distinguishes sharply extracted fringes from the poor ones. To overcome this limitation we proposed in [80] the automatic SR (ASR) method, which analyzes the fringe pattern pixel-by-pixel (whereas the SR used subjective thresholding for the whole region separation). The ASR calculates the intensity modulation distribution for all informative BIMFs (automatically neglecting the last one as a residue and first one in case of noisy interferograms) and takes into the reconstruction process for each pixel the value from the BIMF corresponding to the highest "local" modulation.…”

“…Utilizing unique features of the empirical mode decomposition, namely its adaptive and local ability to separate component structures, the ASR-EFEMD robustness to noise, fringe shape, period and orientation changes, as well as significant variations of contrast and background term is provided. To further advance the ASR-EFEMD algorithm we proposed the Mutual Information Detrending [80] procedure for automatic background term identification in cases when it spreads out of the residue and onto few last BIMFs. It is based on mutual information calculation between each two consecutive BIMFs in search for the drastic drop of similarity between them encountered when fringes disappear.…”

“…Proposed algorithm consists of three main computational routines: (1) Orthogonal Empirical Mode Decomposition employed for the orthogonal fringe set separation, (2) ASR-EFEMD [79,80] enriched by the Mutual Information Detrending procedure applied for further single-family pattern fast and adaptive filtering (bias term and noise removal) and (3) the Hilbert spiral transform for enhanced fringe phase demodulation. HS is used under the assumption of uniform fringe direction which is ensured in the crossed fringe pattern encountered in grating (moiré) interferometry or Talbot interferometry.…”

Fusion of infrared and visible images based on BEMD and NSDFB

“…The EMD is a highly efficient and adaptive method, and offers higher frequency resolution and more accurate timing of nonlinear and non-stationary signal than conventional multi-scale transform techniques [21][22][23]. As EMD has many advantages, the researchers have further extended it to complex empirical mode decomposition (CEMD) [24], bidimensional empirical mode decomposition (BEMD) [25]and multivariate empirical mode decomposition (MEMD) [26], which have been widely used in signal denoising, signal segmentation, signal fusion and so on [21][22][23][24][25][27][28][29]. In this paper, we focus on the EMD in image fusion domain.…”