Authors: Aafreen Alam, Dr. Avinash Rai
DOI Link: https://doi.org/10.22214/ijraset.2022.46014
Certificate: View Certificate
Images produced by synthetic aperture radar (SAR) are crucial for observing and visualizing situations. However, speckle noise makes it difficult to assess SAR images since it reduces image quality and leads to incorrect interpretation. Multiplicative noise features can be found in speckle noise. For the past few years, experts have concentrated on despeckling or speckle reduction. However, the majority of the current efforts showed a loss of edge information. Since wavelet transform and bivariate shrinkage functions have many advantages, this study is devoted to designing a method for speckle removal. After performing a logarithmic transformation to turn multiplicative noise into additive noise, the suggested approach next applies a Lee filter. Then, a wavelet transform was used to breakdown the filtered image. Prior to applying the median filter, the bivariate shrinkage function was used to estimate each coefficient. The simulation results demonstrate that the suggested approach outperforms previous work and several traditional methods.
Synthetic aperture radar (SAR) sensors provide a number of advantages over optical remote sensing, the most significant of which is the ability to record throughout the day and throughout the year . The existence of speckle noise, a type of unwanted or undesirable alteration signal-related granular noise, is the main drawback of SAR images . Over the past three decades, a variety of SAR image de-noising approaches have been put forth. To address the problem, a number of researchers average a fixed number of various photographs but value a considerable loss in picture resolution. The additive model produced by the logarithmic transformation that was initially employed to minimise speckle noise is easier to utilize. In order for certain well-known methods for eliminating distortion to also operate with the modified model, additive white Gaussian noise (AWGN) may be used as a model . Such methods usually ignore a few basic speckle features despite how simple they are to use. The zero mean Normal Distribution, or what is commonly referred to as the Gaussian distribution, is not exactly followed by the log-transformed speckle interference. Therefore, the variance needs to be corrected before continuing the process . During the same time period, de-noising inside this initial domain was tackled by extremely sophisticated algorithms based on the multiplicative speckle paradigm. Such studies unequivocally established the necessity of a certain kind of local adaptation to account for the non-stationary of this image. Additional methods for eliminating distortion in the transform domain are becoming available with the development and improvement of such multi-scale analysis framework. After a homomorphic filtering, wavelet shrinkage might be easily added to such altered coefficients. In addition to the spatial domain, wavelet techniques benefit from spatial adaptation while enhancing the image to successfully maintain both image textures and bounds -.
Over the past few years, some researchers have worked very hard on SAR to eliminate speckles, and many methods have been created, such as the Lee filter , the Kuan filter , the Frost filter , and the maximum a posteriori (MAP) filter . On the other hand, traditional spatial domain techniques frequently over smooth details like corners and texturing, which can occasionally degrade the spatial quality of images. Such filtering techniques are simple and uncomplicated, but they do not preserve visual characteristics like brightness, strength, edges, borders, etc., and system performance is likely to be affected by the relevant terrain. Consequently, transform domain (mathematical method) filters have been developed recently and have produced excellent results. Examples include the wavelet transform , , curvelet transform , , and shearlet transform , . Comparatively, while transform domain techniques effectively minimise speckle, they also have the potential to cause pixel distortion and spurious defects, as well as errors in the preservation of back scatter and detailed information in specific places. Despite the image's valuable local or global features, this is primarily due to the transform domain's inherent inefficiency.
This paper suggests a wavelet-based bivariate shrinking technique that significantly reduces speckle in order to produce high resolution SAR images.
Wavelet transformation can effectively define functions or signals with localised features due to the limited support of wavelet basis functions. Then, using a wavelet-based method, we reduce speckle noise. Last but not least, we found that wavelet-based denoising works better than conventional speckle filters.
This research paper is the result of guidance and support of various people at University Institute of Technology, Rajiv Gandhi Proudyogiki Vishwavidyalaya Bhopal, without whom all my effort would have been directionless and fruitless. I sincerely thank all of them, for supporting me in completing the thesis work.
I express my ardent and earnest gratitude to my guide Dr. Avinash Rai, Department of Electronics & Communication Engineering, University Institute of Technology – Rajiv Gandhi Proudyogiki Vishwavidyalaya, Bhopal for help and Encouragement at all the stages of my thesis work.
I express my heartfelt and profound gratitude to Prof Dr. Sanjay K. Sharma, Head Of Department Electronics & Communication Engineering, University Institute of Technology – Rajiv Gandhi Proudyogiki Vishwavidyalaya, Bhopal for help and Encouragement at all the stages of my thesis work.
I express my heartfelt and profound gratitude to Dr. Sudhir Singh Bhadauria (Director), University Institute of Technology – Rajiv Gandhi Proudyogiki Vishwavidyalaya, Bhopal for help and Encouragement at all the stages of my thesis work.
Finally, I would like to say that I am indebted to my parents and friends for everything that they have done for me. All of this would have been impossible without their constant support. I also thanks to God for being kind to me and driving me through this journey.
In this study, the wavelet transform is used to scale the image, eliminate speckle noise, and create a multi-resolution representation. The paper uses a bivariate shrinkage function in combination with wavelet decomposition at various noise variance levels to achieve this. The outcome is assessed based on various performance metrics and contrasted with some already in use strategies. Results allow for the development of more efficient approach. Future research can expand on this work by using dual tree complex wavelet transforms to denoise and de-blur multiple resolution images.
 Achim, A., Tsakalides, P., &Bezerianos, A. (2003). SAR image denoising via Bayesian wavelet shrinkage based on heavy-tailed modeling. IEEE Transactions on Geoscience and Remote Sensing, 41(8), 1773–1784. https://doi.org/10.1109/TGRS.2003.813488.  Zhao, Q., Wu, S., Zhang, Z., Sun, Y., Fu, Y., & Wang, H. (2021). SAR Image Noise Reduction Based on Wavelet Transform and DnCNN. 1–4. https://doi.org/10.23919/CISS51089.2021.9652335.  Argenti, F., Bianchi, T., &Alparone, L. (2006). Multiresolution MAP despeckling of SAR images based on locally adaptive generalized Gaussian pdf modeling. IEEE Transactions on Image Processing, 15(11), 3385–3399. https://doi.org/10.1109/TIP.2006.881970.  Castaneda, R., Garcia-Sucerquia, J., &Doblas, A. (2021). Speckle noise reduction in coherent imaging systems via hybrid median–mean filter. https://doi.org/10.1117/1.OE.60.12.123107.  Chen, G., Li, G., Liu, Y., Zhang, X. P., & Zhang, L. (2018). SAR image despeckling by combination of fractional-order total variation and nonlocal low rank regularization. Proceedings - International Conference on Image Processing, ICIP, 2017-September, 3210–3214. https://doi.org/10.1109/ICIP.2017.8296875.  Choi, H., &Jeong, J. (2018). Despeckling images using a preprocessing filter and discrete wavelet transform-based noise reduction techniques. IEEE Sensors Journal, 18(8), 3131–3139. https://doi.org/10.1109/JSEN.2018.2794550.  Frost, V. S., Stiles, J. A., Shanmugan, K. S., & Holtzman, J. C. (1982). A Model for Radar Images and Its Application to Adaptive Digital Filtering of Multiplicative Noise. IEEE Transactions on Pattern Analysis and Machine Intelligence, PAMI-4(2), 157–166. https://doi.org/10.1109/TPAMI.1982.4767223.  Gleich, D., Kseneman, M., &Datcu, M. (2010). Despeckling of terraSAR-X data using second-generation wavelets. IEEE Geoscience and Remote Sensing Letters, 7(1), 68–72. https://doi.org/10.1109/LGRS.2009.2020610.  Guan, D., Xiang, D., Hu, C., & Zhong, Z. (2019). Sar Image Despeckling with the Multi-Scale Nonlocal Low-Rank Model. International Geoscience and Remote Sensing Symposium (IGARSS), 2941–2944. https://doi.org/10.1109/IGARSS.2019.8899791.  Guan, D., Xiang, D., Tang, X., &Kuang, G. (2019). SAR Image Despeckling Based on Nonlocal Low-Rank Regularization. IEEE Transactions on Geoscience and Remote Sensing, 57(6), 3472–3489. https://doi.org/10.1109/TGRS.2018.2885089.  Geng, J., Fan, J., Ma, X., Wang, H., & Cao, K. (2016). An iterative low-rank representation for SAR image despeckling. International Geoscience and Remote Sensing Symposium (IGARSS), 2016-November, 72–75. https://doi.org/10.1109/IGARSS.2016.7729009.  Hou, B., Zhang, X., Bu, X., & Feng, H. (2012). SAR image despeckling based on nonsubsampled shearlet transform. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 5(3), 809–823. https://doi.org/10.1109/JSTARS.2012.2196680.  Jain, L., & Singh, P. (2020). A novel wavelet thresholding rule for speckle reduction from ultrasound images. Journal of King Saud University - Computer and Information Sciences. https://doi.org/10.1016/J.JKSUCI.2020.10.009  Jain, V., Shitole, S., Turkar, V., & Das, A. (2020). Impact of DFT based speckle reduction filter on classification accuracy of synthetic aperture radar images. 2020 IEEE India Geoscience and Remote Sensing Symposium, InGARSS 2020 - Proceedings, 54–57. https://doi.org/10.1109/INGARSS48198.2020.9358943.  Khare, S., & Kaushik, P. (2021). Speckle filtering of ultrasonic images using weighted nuclear norm minimization in wavelet domain. Biomedical Signal Processing and Control, 70, 102997. https://doi.org/10.1016/J.BSPC.2021.102997.  Ko, J., & Lee, S. (2022). SAR Image Despeckling Using Continuous Attention Module. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 15, 3–19. https://doi.org/10.1109/JSTARS.2021.3132027.  Kuan, D. T., Sawchuk, A. A., Strand, T. C., &Chavel, P. (1985). ADAPTIVE NOISE SMOOTHING FILTER FOR IMAGES WITH SIGNAL-DEPENDENT NOISE. IEEE Transactions on Pattern Analysis and Machine Intelligence, PAMI-7(2), 165–177. https://doi.org/10.1109/TPAMI.1985.4767641.  Lee, J. Sen. (1980). Digital Image Enhancement and Noise Filtering by Use of Local Statistics. IEEE Transactions on Pattern Analysis and Machine Intelligence, PAMI-2(2), 165–168. https://doi.org/10.1109/TPAMI.1980.4766994.  Liu, S., Shi, M., Hu, S., & Xiao, Y. (2014). Synthetic aperture radar image de-noising based on Shearlet transform using the context-based model. Physical Communication, 13(PC), 221–229. https://doi.org/10.1016/J.PHYCOM.2014.02.002.  Rahimizadeh, N., Hasanzadeh, R. P., &Janabi-Sharifi, F. (2020). An optimized non-local LMMSE approach for speckle noise reduction of medical ultrasound images. Multimedia Tools and Applications 2020 80:6, 80(6), 9231–9253. https://doi.org/10.1007/S11042-020-10051-Z.  Ranjani, J. J., &Thiruvengadam, S. J. (2010). Dual-tree complex wavelet transform based sar despeckling using interscale dependence. IEEE Transactions on Geoscience and Remote Sensing, 48(6), 2723–2731. https://doi.org/10.1109/TGRS.2010.2041241.  Rebar Ihsan, R., Almufti, S. M., &Marqas, R. B. (2020). A Median Filter With Evaluating of Temporal Ultrasound Image for Impulse Noise Removal for Kidney Diagnosis. Journal of Applied Science and Technology Trends, 1(2), 71–77. https://doi.org/10.38094/jastt1217.  Shang, R., Liu, M., Lin, J., Feng, J., Li, Y., Stolkin, R., & Jiao, L. (2022). SAR Image Segmentation Based on Constrained Smoothing and Hierarchical Label Correction. IEEE Transactions on Geoscience and Remote Sensing, 60. https://doi.org/10.1109/TGRS.2021.3076446.  Singh, P., Shree, R., & Diwakar, M. (2021). A new SAR image despeckling using correlation based fusion and method noise thresholding. Journal of King Saud University - Computer and Information Sciences, 33(3), 313–328. https://doi.org/10.1016/J.JKSUCI.2018.03.009.  Sujitha, A. G., Vasuki, D. P., & Deepan, A. A. (2019). Hybrid Laplacian Gaussian Based Speckle Removal in SAR Image Processing. Journal of Medical Systems, 43(7). https://doi.org/10.1007/S10916-019-1299-0.  Wang, B. H., Zhao, C. Y., & Liu, Y. Y. (2018). An improved SAR interferogram denoising method based on principal component analysis and the Goldstein filter. Remote Sensing Letters, 9(1), 81–90. https://doi.org/10.1080/2150704X.2017.1392633  Yahia, M., Ali, T., Mortula, M. M., Abdelfattah, R., Mahdy, S. El, &Arampola, N. S. (2020). Enhancement of SAR Speckle Denoising Using the Improved Iterative Filter. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 13, 859–871. https://doi.org/10.1109/JSTARS.2020.2973920  Yan, P., Chen, C., & Chen, C. (2005). An algorithm for filtering multiplicative noise in wide range Un algorithme pour le filtrage du bruit multiplicatif APPLICATIONS.
Copyright © 2022 Aafreen Alam, Dr. Avinash Rai. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Paper Id : IJRASET46014
Publish Date : 2022-07-26
ISSN : 2321-9653
Publisher Name : IJRASET
DOI Link : Click Here