|
|
基于深度先验的乘性噪声图像去噪
|
Abstract:
深度图像先验作为典型的无监督学习方法，不需要采用大量的训练样本，因此在一些难以采集大量训练样本训练数据集的领域有许多应用前景，另外，采集大量训练样本需要耗费大量的人力物力，并且可能出现模型的过拟合问题。在本文中，为了有效地去除图像中的乘性噪声，提出了一种基于深度先验的图像去噪变分模型。为求解新模型，本文采用交替方向法求解该问题。数值实验表明，相比其它先进的方法，本文的模型取得更好的去噪结果，具有更好的图像视觉效果。
As a typical unsupervised learning method, deep image prior does not require a large number of training samples, so it has many application prospects in some fields where it is difficult to collect a large number of training samples and training data sets. In addition, collecting a large number of training samples requires a lot of manpower and material resources, and the problem of overfitting of the model may occur (and the model may be over-fitted). In this paper, to effectively remove multiplicative noise in images, a deep prior-based variational model for image denoising is pro-posed. To solve the new model, this paper adopts the alternating direction method to solve the problem. Numerical experiments show that the proposed model in this paper can remove multipli-cative noise more effectively than other state-of-the-art methods, and has better image visual ef-fects.
| [1] | 阮秋琦. 数字图像处理学[M]. 北京: 电子工业出版社, 2007. |
| [2] | Aubert, G. and Aujor, J. (2008) A Variational Approach to Removing Multiplicative Noise. SIAM Journal on Applied Mathematics, 68, 925-946. https://doi.org/10.1137/060671814 |
| [3] | Shi, J. and Osher, S. (2008) A Nonlinear Inverse Scale Space Method for a Convex Multiplicative Noise Model. SIAM Journal on Imaging Sciences, 1, 294-321. https://doi.org/10.1137/070689954 |
| [4] | Huang, Y., Ng, M. and Wen, Y. (2009) A New Total Variation Method for Multiplicative Noise Removal. SIAM Journal on Imaging Sciences, 2, 20-40. https://doi.org/10.1137/080712593 |
| [5] | Hu, S., Lin, G. and Lu, C. (2021) GPX-ADMM-Net: ADMM-Based Neural Network with Generalized Proximal Operator. 2020 28th European Signal Processing Conference (EUSIPCO), Amsterdam, 2055-2059.
https://doi.org/10.23919/Eusipco47968.2020.9287399 |
| [6] | Dittmer, S., Kluth, T., Maass, P. and Baguer, D.O. (2020) Regularization by Architecture: A Deep Prior Approach for Inverse Problems. Journal of Mathematical Imaging and Vision, 62, 456-470.
https://doi.org/10.1007/s10851-019-00923-x |
| [7] | Zhang, K., Zuo, K., Chen, K., Meng, K. and Zhang, K. (2017) Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising. IEEE Transactions on Image Pro-cessing, 26, 3142-3155.
https://doi.org/10.1109/TIP.2017.2662206 |
| [8] | Zhang, K., Zuo, K. and Zhang, K. (2018) Ffdnet: Toward a Fast and Flexible Solution for CNN-Based Image Denoising. IEEE Transactions on Image Processing, 27, 4608-4622. https://doi.org/10.1109/TIP.2018.2839891 |
| [9] | Xu, W., Zhao, X., Jiang, T., Wang, Y. and Ng, M.K. (2019) Deep Plug-and-Play Prior for Low-Rank Tensor Completion. Neurocomputing, 400, 137-149. https://doi.org/10.1016/j.neucom.2020.03.018 |
| [10] | Ulyanov, D., Vedaldi, A. and Lempitsky, V. (2020) Deep Im-age Prior. International Journal of Computer Vision, 128, 1867-1888. https://doi.org/10.1007/s11263-020-01303-4 |
| [11] | Cascarano, P., Sebastiani, A., Comes, M., Franchini, G. and Porta, F. (2020) Combining Weighted Total Variation and Deep Image Prior for Natural and Medical Image Restoration via ADMM. 2021 21st International Conference on Computational Science and Its Applications (ICCSA), 39-46. https://doi.org/10.1109/ICCSA54496.2021.00016 |
| [12] | Long, J., Shelhamer, E. and Darrell, T. (2015) Fully Con-volutional Networks for Semantic Segmentation. CVPR, 3431- 3440. https://doi.org/10.1109/CVPR.2015.7298965 |
| [13] | Rudin, L., Osher, S. and Fatemi, E. (1992) Nonlinear Total Variation Based Noise Removal Algorithms. Physical D, 60, 259-268. https://doi.org/10.1016/0167-2789(92)90242-F |
