English   español  
Please use this identifier to cite or link to this item: http://hdl.handle.net/10261/11608
logo share SHARE   Add this article to your Mendeley library MendeleyBASE

Visualizar otros formatos: MARC | Dublin Core | RDF | ORE | MODS | METS | DIDL
Exportar a otros formatos:

Structural noise reduction using multiresolution-based spectral substraction

AuthorsIzquierdo, M. A. G.; Hernández, M. G.; Molero Armenta, Miguel Ángel; Anaya Velayos, José Javier
KeywordsUltrasonic NDE/NDT
dispersive materials
SNR enhancement
Issue Date16-Mar-2009
AbstractIn ultrasonic non destructive evaluation (UT-NDE) of highly scattering materials, detection of little flaw echoes is difficult due to the masking effect of the structural noise. Most of the proposed techniques for enhancing the flaw visibility do not exploit the spectra of the structural noise and the incoming flaw signal depth dependency. They assume homogeneous behavior of the insonified material in order to design the processing algorithms. In this work, a multiresolution-based spectral subtraction method for enhancing the flaw signal to structural noise ratio (SNR) is proposed. This method, using the wavelet transform, exploits the non-stationary characteristic of the structural noise. First step of this algorithm is the time-scale estimation of several statistics of the structural noise. On the second step, non-linear over-subtraction of the raw signal wavelet transform is applied. This strategy produces improved results in SNR compared with linear and non- over-subtraction techniques. Next, a post-processing based on exploiting additional information from the physical knowledge of the structural noise is applied and finally the inverse Wavelet Transform obtains the output processed signal. Experimental results are also included showing that the proposed method dramatically increase the visibility of the flaw echoes.
Publisher version (URL)http://dx.doi.org/10.3728/ICUltrasonics.2007.Vienna.1290_izquierdo
Appears in Collections:(IAI) Comunicaciones congresos
Files in This Item:
File Description SizeFormat 
1290_izquierdo.pdf281,73 kBAdobe PDFThumbnail
Show full item record
Review this work

WARNING: Items in Digital.CSIC are protected by copyright, with all rights reserved, unless otherwise indicated.