**Part 4**

**Image Processing** 

370 Advances in Wavelet Theory and Their Applications in Engineering, Physics and Technology

[18] Newland, E., 1994, Wavelet Analysis, Part II: Wavelet Maps, *Journal of Sound and* 

[20] Lin, J., Zuo, M., J., (2003), Gearbox fault diagnosis using adaptive wavelet filter,

[21] Boulahbal, D., Golnaraghi M., F., Ismail, F., (1999), Amplitude and phase wavelet maps

[22] Baydar, N., Ball, A., (2003), Detection of gear failures via vibration and acoustic signals using wavelet transform, *Mechanical Systems and Signal Processing* 17(4), 787-804. [23] Meltzer, G., Dien, N., P., (2004), Fault diagnosis in gears operating under non-

[24] Wang, W., J., (1995), Application of orthogonal wavelets to early gear damage

[25] Berri, S., Klosner, J., M., (1999), A new strategy for detecting gear faults using

[26] Wang, W. J., McFadden, P. D., (1996), Application of wavelets to gearbox vibration signals for fault detection, *Journal of Sound and Vibration*, 192, 927–939, 1996. [27] D' Elia, G., 2008, Ph.D. Thesis in Applied Machines, Fault detection in rotating machines by vibration signal processing techniques, Universita' di Bologna, Italy. [28] Schukin, E.L., Zamaraev, R.U., Schukin, L.I., (2004), The optimization of wavelet

[29] Yang, W., (2007), A natural way for improving the accuracy of the continuous wavelet

[30] Halim B. et al., (2008), Time domain averaging across all scales: A novel method for

[31] Addison P. S., (2002), The Illustrated Wavelet Transform Handbook, Istitute of Physics

[32] Delvecchio, S., D'Elia, G., Mucchi, E. and Dalpiaz, G. (2010). Advanced Signal

transform. *Journal of Sound and Vibration*, 306, 928-939.

detection, *Mechanical Systems and Signal Processing* 9(5), 497-507.

for the detection of cracks in geared systems, *Mechanical Systems and Signal* 

stationary rotational speed using polar wavelet amplitude maps, *Mechanical* 

denoising with the orthogonal Discrete Wavelet Transform (ODWT), in *Proceedings of the 1999 ASME Design Engineering Technical Conferences*, September 12-15, 1999,

transform for the impulse analysis in vibration signals. *Mechanical Systems and* 

detection of gearbox faults, *Mechanical Systems and Signal Processing*, 22, pp. 261-

Processing Tools for the Vibratory Surveillance of Assembly Faults in Diesel Engine Cold Tests. *ASME Journal of Vibration and Acoustics*, Volume 132, Issue 2,

[19] Mallat, S., A wavelet tour of signal processing. Academic Press, 1999.

*Mechanical Systems and Signal Processing* 17(6), 1259-1269.

*Vibration* 116, 417-425.

*Processing* 13(3), 423-436.

Las Vegas, Nevada.

278.

*Signal Processing*, 18, 1315-1333.

Publishing, Philadelphia.

021008 (10 pages).

*Systems and Signal Processing* 18, 985-992.

**16** 

**Information Extraction and** 

Matej Kseneman1 and Dušan Gleich2

*1Margento R&D d.o.o.,* 

*Slovenia* 

**Despeckling of SAR Images with** 

**Second Generation of Wavelet Transform** 

*2University of Maribor, Faculty of Electrical Engineering and Computer Science,* 

Synthetic Aperture Radar (SAR) technology is mainly used to obtain high-resolution images of ground areas in resolutions even less than meter. SAR is even capable of imaging a wide area of terrain and from two and more images it is possible to reconstruct a 3D digital elevation model of ground terrain. Good thing about SAR is an all whether operation and possibility to capture images under various inclination angles. Because digital images are usually corrupted by noise that arises from an imaging device, there is always a need for a good filtering algorithm to remove all disturbances, thus enabling more information extraction. The SAR images are corrupted by a noise called speckle, which makes the interpretation of SAR images very difficult. The goal of removing speckles from the SAR image is to represent a noise-free image and preserve all important features of the SAR

Many different techniques for SAR image despeckling have been proposed over the past few years. Speckle is a noise-like characteristic of SAR images and it is a multiplicative nature, if the intensity or amplitude image is observed. The despeckling can be performed in the image or in the frequency domain. The well-known despeckling filters are Lee (Lee, 1980), Kuan (Kuan et al., 1985), and Frost (Frost et al., 1982). Lee and Kuan filters can be considered as an adaptive mean filters, meanwhile the Frost filter can be considered as a mean adaptive weighted filter. The Bayesian filters are based on the Bayesian theorem, which defines a posterior probability by using a prior, likelihood and evidence probability density functions (pdf). The solution for noise-free image is found by a maximum a posteriori (MAP) estimate. The MAP estimate of a noise free image was proposed in (Walessa & Datcu, 2000), where the noise free image was approximated by a Gauss-Markov random field prior and the noise was modeled with Gamma pdf. Model based despeckling and information extraction is one of the promising techniques of SAR image denoising and scene interpretation. The wavelet based despeckling algorithms have been proposed in (Dai et al., 2004), (Argenti & Alparone, 2002), and (Foucher et al., 2001). The second generation wavelets Chirplet (Cui & Wong, 2006), Contourlet (Chuna et al., 2006), Bandelet (Le Pennec

image, as for example edges, textures, region borders, etc.

& Mallat, Apr 2005) have appeared over the past few years.

**1. Introduction** 
