Open Access Open Access  Restricted Access Subscription or Fee Access

WNNM with WMF for Improvement in Image Denoising and Image Segmentation Accuracy

Sicong Li

Abstract


Denoising is an important task in image processing and diversity of algorithms have been proposed in image denoising. Weighted nuclear norm minimization (WNNM) is one of them. Despite good performance WNNM has in removing non-sparse noise, it is not so powerful in sparse noise denoising like salt-pepper noise. This study proposes a novel method combining weighted median filter (WMF) with WNNM to improve the denoising effect. Furthermore, this research also implements Markov random filed (MRF) to verify this improvement in denoising effect which can effectively improve the accuracy in image segmentation. Within our experimental results, PSNR result increases by 10.04 dB in maximum with WMF being added. Besides, the results show that our proposed method has better performance than combining traditional median filter with WNNM in most situations and it is better than combination of adaptive median filter and WNNM when salt-pepper noise is small. Finally, the experimental results show that our method is more effective than WNNM with more iterations in achieving higher PSNR result because time consumed in our method is remarkably less.


Full Text:

PDF

References


Elad M, Aharon M. Image denoising via sparse and redundant representations over learned dictionaries. IEEE Trans Image Process. 2006 Dec; 15(12): 3736–3745.

Dabov K, Foi A, Katkovnik V, Egiazarian K. Image denoising by sparse 3-d transform-domain collaborative filtering. IEEE Trans Image Process. 2007 Aug; 16(8): 2080–2095.

Coll Buads B, Morel J-M. A non-local algorithm for image denoising. In Proc IEEE CVPR. 2005; 2: 60–65.

Gu S, Zhang L, Zuo W, Feng X. Weighted Nuclear Norm Minimization with Application to Image Denoising. IEEE CVPR. 2014; 2682–2869.

Dabov K, Foi A, Katkovnik V, Egiazarian K. BM3D Image Denoising with Shape-Adaptive Principal Component Analysis. Signal Processing with Adaptive Sparse Structured Representations, Inria Rennes - Bretagne Atlantique, Saint Malo, France. 2009 Apr.

Romano Y, Elad M. Improving K-SVD denoising by post-processing its method-noise. 2013 IEEE International Conference on Image Processing. 2013; 435–439.

Guyon G, Bouwmans T, Zahzah E. Robust principal component analysis for background subtraction: Systematic evaluation and comparative analysis. In: Principal Component Analysis. 2012; 223–228.

Javed S, Oh S, Sobral A, Bouwmans T, Jung S. OR-PCA with MRF for Robust Foreground Detection in Highly Dynamic Backgrounds. ACCV 2014. 2015; 284–299.

Wu J, Lee X. An Improved WNNM Algorithm for Image Denoising. ICSP 2019. 2019; 1237(2): 1–6.

Candes EJ, Recht B. Exact Matrix Completion via Convex Optimization. Found Comput Math. 2009; 9(6): 717–772.

Hwang H, Hadded RA. Adaptive median filter: New algorithms and results. IEEE Trans Image Process. 1995; 4(4): 499–502.

Lu C, Chou T. Denoising of salt-and-pepper noise corrupted image using modified directionalweighted-median filter. Pattern Recognit Lett. 2012; 33: 1287–1295.

Li T, Tang S, Wang F, Tong M, Xu C. Image Enhancement Study Based on Adaptive Median Filtering with Secondary Noise Detection and Neighborhood Pixel Recovery. 2018 International Conference on Robots & Intelligent System (ICRIS). 2018; 134–136.

Yang R, Yin L, Gabbouj M, Neuvo Y. Circuits and Systems Exposition Weighted Median Filters: A Tutorial. IEEE Trans Circuits Syst II Analog Digit Signal Process. 1996 Mar; 43(3): 157–192.

Ahmadvand, Yousefi S. A novel Markov random field model based on region adjacency graph for T1 magnetic resonance imaging brain segmentation. Int J Imaging Syst Technol. 2017 Mar; 27(1):

–88.


Refbacks

  • There are currently no refbacks.