Noise removal refers to the most vital process in image processing to remove the noise from a given image and reconstruct the original image. Among many denoising techniques, four types of extended versions of robust Annihilating filter-based Low-rank Hankel Matrix (r-ALOHA) approaches have been proposed in the previous researches. In those approaches, different kinds of transform domains like log-exponential, wavelet, generalized Hough, and gradient were considered separately in which that the image patch was considered as it was sparse in the considered transform domains independently to denoise the corrupted image. Even if gradient transform based denoising called e4-ALOHA achieves better performance than the other transform domains, it requires an automatic selection of Optimal Patch Size (OPS) to further improve the denoising performance. Hence in this article, an automatic selection of OPS is proposed with e4-ALOHA that searches similar image patches and selects an optimal patch size. In this technique, a Flower Pollination optimization Algorithm (FPA) is proposed to search similar patches and choose an optimal patch size adaptively according to the variance of similar patch groups. Once an optimal patch size is selected, e4-ALOHA is applied to perform the denoising process. Finally, the effectiveness of the proposed technique is evaluated through the experimental results.

Noise removal, r-ALOHA, e4-ALOHA, Optimal patch size, Flower pollination algorithm Formatting

[1] P. Kamboj, V. Rani, “A brief study on various noise model and filtering techniques”, Journal of Global Research in Computer Science, Vol.4, Issue.4, pp.166-171, 2013.
[2] A. Suganthi, M. Senthilmurugan, “Comparative study of various impulse noise reduction techniques”, International Journal of Engineering Research and Applications, Vol.3, Issue.5, pp.1302-1306, 2013.
[3] K. Pritamdas, Kh. M. Singh, L. L. Singh, “A summary on various impulse noise removal techniques”, International Journal of Science and Research, Vol.6, Issue.3, pp.941-954, 2017.
[4] K. H. Jin, J. C. Ye, “Sparse and low-rank decomposition of a hankel structured matrix for impulse noise removal”, IEEE Transactions on Image Processing, Vol.27, Issue.3, pp.1448-1461, 2018.
[5] L. B. Victoria, S. Sathappan, “A sporadic decomposition of hankel structured matrix in logarithmic and wavelet domain for impulse noise removal”, International Journal of Engineering & Technology, 2018.
[6] L. B. Victoria, S. Sathappan, “A sporadic decomposition of hankel structured matrix in generalized Hough and gradient transform domain for impulse noise removal”, In International Conference on Research Trends in Computing Technologies (ICRTCT18), 2018.
[7] C. A. Deledalle, J. Salmon, A. S. Dalalyan, “Image denoising with patch based PCA: local versus global”, In BMVC, Vol.81, pp.425-455, 2011.
[8] X. Zhang, X. Feng, W. Wang, “Two-direction nonlocal model for image denoising”, IEEE Transactions on Image Processing, Vol.22, Issue.1, pp.408-412, 2013.
[9] X. Chen, S. Bing Kang, J. Yang, J. Yu, “Fast patch-based denoising using approximated patch geodesic paths”, In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp.1211-1218, 2013.
[10] C. Kervrann, “PEWA: Patch-based exponentially weighted aggregation for image denoising”, In Advances in Neural Information Processing Systems, pp.2150-2158, 2014.
[11] K. Kundu, “Image Denoising using Patch based Processing with Fuzzy Gaussian Membership Function”, International Journal of Computer Applications, Vol.118, Issue.12, pp.35-40, 2015.
[12] X. Liu, Z. Yang, J. Wang, J. Liu, K. Zhang, W. Hu, “Patch-based denoising method using low-rank technique and targeted database for optical coherence tomography image”, Journal of Medical Imaging, Vol.4, Issue.1, pp.014002(1-12), 2017.
[13] M. Kaur, B. Jindal, “Improved sparse matrix denoising technqiues using affinity matrix for geographical images”, International Journal of Scientific Research in Computer Science and Engineering, Vol.5, Issue.5, pp.51-57, 2017.