双路融合注意力机制的自适应图像去噪算法
Dual-Path Fusion Attention Mechanism for Adaptive Image Denoising Algorithm
DOI: 10.12677/pm.2026.163081, PDF,    科研立项经费支持
作者: 王潇旖, 仲彦军*, 资 政：新疆师范大学数学科学学院，新疆 乌鲁木齐；新疆师范大学CAD & CG实验室，新疆 乌鲁木齐
关键词: 深度学习；图像去噪；动态加权；通道注意力；多尺度；Deep Learning； Image Denoising； Dynamic Weighting； Channel Attention； Multi-Scale
摘要: 针对现有图像去噪算法在特征处理阶段存在的不足，如未能充分利用局部特征、难以有效恢复边缘细节，甚至可能引起图像失真等问题，本文提出了一种结合传统去噪和深度学习的去噪算法，旨在提高去噪效果并保留图像细节和边缘。首先，使用自适应权重滤波根据噪声水平动态调整双边滤波和非局部均值滤波的权重，以平衡去噪与细节保留；其次，通过不同尺寸的卷积核提取图像的局部到全局特征；最后，应用通道注意力机制强化关键特征，提高模型对多尺度特征的感知能力。实验结果证明，与当前流行的图像去噪技术相比，该方法不仅能够显著降低图像中的噪声，还能够恢复出更精细的纹理效果，在定性和定量分析中均表现出优异的去噪性能。
Abstract: Existing image denoising methods often fail to fully utilize local features, effectively restore edge details, and may even introduce image distortion. This paper proposes a denoising algorithm that integrates traditional methods with deep learning to enhance denoising performance while preserving image details and edges. The model employs adaptive weight filtering to dynamically balance bilateral filtering and non-local means filtering according to noise levels. Multiple convolutional kernels extract local-to-global features, while a channel attention mechanism enhances critical features and improves multi-scale feature perception. Experimental results demonstrate that the proposed method achieves superior noise reduction and recovers finer textures compared to current denoising techniques, exhibiting excellent performance in both qualitative and quantitative evaluations.
文章引用：王潇旖, 仲彦军, 资政. 双路融合注意力机制的自适应图像去噪算法[J]. 理论数学, 2026, 16(3): 177-190. https://doi.org/10.12677/pm.2026.163081

参考文献

[1] Piao, W.Y., Yuan, Y.B. and Lin, H.J. (2018) A Digital Image Denoising Algorithm Based on Gaussian Filtering and Bilateral Filtering. ITM Web of Conferences, 17, Article No. 01006. [Google Scholar] [CrossRef] 
[2] Yao, C., Jin, S., Liu, M. and Ban, X. (2022) Dense Residual Transformer for Image Denoising. Electronics, 11, Article No. 418. [Google Scholar] [CrossRef] 
[3] Gao, Y.B., Wen, Z.C., Duan, X.S., et al. (2024) Image Denoising Method Based on Multi-Scale Dual-Branch Transformer. Journal of Yunnan University: Natural Sciences Edition, 47, 465-474.
[4] Jia, X., Liu, S., Feng, X. and Zhang, L. (2019) Focnet: A Fractional Optimal Control Network for Image Denoising. 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Long Beach, 16-20 June 2019, 6054-6063. [Google Scholar] [CrossRef] 
[5] Dabov, K., Foi, A., Katkovnik, V. and Egiazarian, K. (2007) Image Denoising by Sparse 3-D Transform-Domain Collaborative Filtering. IEEE Transactions on Image Processing, 16, 2080-2095. [Google Scholar] [CrossRef] [PubMed]
[6] Zhang, K., Zuo, W., Chen, Y., Meng, D. and Zhang, L. (2017) Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising. IEEE Transactions on Image Processing, 26, 3142-3155. [Google Scholar] [CrossRef] [PubMed]
[7] Zhang, K., Zuo, W. and Zhang, L. (2018) FFDNet: Toward a Fast and Flexible Solution for CNN-Based Image Denoising. IEEE Transactions on Image Processing, 27, 4608-4622. [Google Scholar] [CrossRef] [PubMed]
[8] Guo, S., Yan, Z., Zhang, K., Zuo, W. and Zhang, L. (2019) Toward Convolutional Blind Denoising of Real Photographs. 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Long Beach, 16-21 June 2019, 1712-1722. [Google Scholar] [CrossRef] 
[9] Anwar, S. and Barnes, N. (2019) Real Image Denoising with Feature Attention. 2019 IEEE/CVF International Conference on Computer Vision (ICCV), Seoul, 27-28 October 2019, 3155-3164. [Google Scholar] [CrossRef] 
[10] Tassano, M., Delon, J. and Veit, T. (2019) DVDNET: A Fast Network for Deep Video Denoising. 2019 IEEE International Conference on Image Processing (ICIP), Taipei, 22-25 September 2019, 1805-1809. [Google Scholar] [CrossRef] 
[11] Sara, U., Akter, M. and Uddin, M.S. (2019) Image Quality Assessment through FSIM, SSIM, MSE and PSNR—A Comparative Study. Journal of Computer and Communications, 7, 8-18. [Google Scholar] [CrossRef] 
[12] Gong, Y., Yu, X., Ding, Y., Peng, X., Zhao, J. and Han, Z. (2021) Effective Fusion Factor in FPN for Tiny Object Detection. 2021 IEEE Winter Conference on Applications of Computer Vision (WACV), 5-9 January 2021, 1160-1168. [Google Scholar] [CrossRef] 
[13] Zhu, P., Li, Z. and Ogino, M. (2023) Recurrent Continuous Adversarial Balancing for Estimating Individual Treatment Effect over Time. International Journal of Computational Intelligence Systems, 16, Article No. 82. [Google Scholar] [CrossRef] 
[14] Ding, X., Guo, Y., Ding, G. and Han, J. (2019) ACNet: Strengthening the Kernel Skeletons for Powerful CNN via Asymmetric Convolution Blocks. 2019 IEEE/CVF International Conference on Computer Vision (ICCV), Seoul, 27 October-2 November 2019, 1911-1920. [Google Scholar] [CrossRef] 
[15] Mairal, J., Bach, F., Ponce, J., Sapiro, G. and Zisserman, A. (2009) Non-Local Sparse Models for Image Restoration. 2009 IEEE 12th International Conference on Computer Vision, Kyoto, 29 September-2 October 2009, 2272-2279. [Google Scholar] [CrossRef] 
[16] Guo, Y., Şengür, A. and Ye, J. (2014) A Novel Image Thresholding Algorithm Based on Neutrosophic Similarity Score. Measurement, 58, 175-186. [Google Scholar] [CrossRef] 
[17] Gu, S., Zhang, L., Zuo, W. and Feng, X. (2014) Weighted Nuclear Norm Minimization with Application to Image Denoising. 2014 IEEE Conference on Computer Vision and Pattern Recognition, Columbus, 24-27 June 2014, 2862-2869. [Google Scholar] [CrossRef] 
[18] Dabov, K., Foi, A., Katkovnik, V. and Egiazarian, K. (2007) Color Image Denoising via Sparse 3D Collaborative Filtering with Grouping Constraint in Luminance-Chrominance Space. 2007 IEEE International Conference on Image Processing, Vol. 1, I-313. [Google Scholar] [CrossRef] 
[19] Johnson, O.V., Xinying, C., Khaw, K.W. and Lee, M.H. (2023) ps-CALR: Periodic-Shift Cosine Annealing Learning Rate for Deep Neural Networks. IEEE Access, 11, 139171-139186. [Google Scholar] [CrossRef] 
[20] Quan, Y., Chen, M., Pang, T. and Ji, H. (2020) Self2self with Dropout: Learning Self-Supervised Denoising from Single Image. 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, 13-19 June 2020, 1890-1898. [Google Scholar] [CrossRef] 
[21] Szegedy, C., Liu, W., Jia, Y.Q., Sermanet, P., Reed, S., Anguelov, D., et al. (2015) Going Deeper with Convolutions. 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Boston, 7-12 June 2015, 1-9. [Google Scholar] [CrossRef] 
[22] Scetbon, M., Elad, M. and Milanfar, P. (2021) Deep K-SVD Denoising. IEEE Transactions on Image Processing, 30, 5944-5955. [Google Scholar] [CrossRef] [PubMed]
[23] Zhang, Y., Li, K., Li, K., Wang, L., Zhong, B. and Fu, Y. (2018) Image Super-Resolution Using Very Deep Residual Channel Attention Networks. In: Ferrari, V., et al., Eds., Computer Vision—ECCV 2018, Springer International Publishing, 294-310. [Google Scholar] [CrossRef] 
[24] Martin, D., Fowlkes, C., Tal, D., et al. (2001) A Database of Human Segmented Natural Images and Its Application to Evaluating Segmentation Algorithms and Measuring Ecological Statistics. Proceedings 8th IEEE International Conference on Computer Vision, Vancouver, 7-14 July 2001, 416-423.
[25] Lee, X.D. and Wu, J.F. (2019) Image Denoising Algorithm Based on Improved NCSR Model. Journal of Physics: Conference Series, 1314, Article ID: 012209. [Google Scholar] [CrossRef] 
[26] Molchanov, P., Mallya, A., Tyree, S., Frosio, I. and Kautz, J. (2019) Importance Estimation for Neural Network Pruning. 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Long Beach, 16-21 June 2019, 11264-11272. [Google Scholar] [CrossRef] 
[27] Wu, X. (2011) Color Demosaicking by Local Directional Interpolation and Nonlocal Adaptive Thresholding. Journal of Electronic Imaging, 20, Article ID: 023016. [Google Scholar] [CrossRef] 
[28] Loshchilov, I. and Hutter, F. (2016) SGDR: Stochastic Gradient Descent with Warm Restarts.
[29] Li, Y.F. and Chen, Y.Y. (2023) NSNet: An N-Shaped Convolutional Neural Network with Multi-Scale Information for Image Denoising. Mathematics, 11, Article No. 2772. [Google Scholar] [CrossRef] 
[30] Zhang, Q., Xiao, J.Y., Zhang, S.C., et al. (2024) Texture-Guided CNN for Image Denoising. Multimedia Tools and Applications, 83, 1-25.
[31] Murali, V. and Sudeep, P.V. (2020) Image Denoising Using DnCNN: An Exploration Study. In: Jayakumari, J., et al., Eds., Advances in Communication Systems and Networks, Springer, 847-859. [Google Scholar] [CrossRef] 
[32] Fang, F.M., Li, J.C., Yuan, Y.T., Zeng, T. and Zhang, G. (2021) Multilevel Edge Features Guided Network for Image Denoising. IEEE Transactions on Neural Networks and Learning Systems, 32, 3956-3970. [Google Scholar] [CrossRef] [PubMed]
[33] Chen, C., Xiong, Z.W., Tian, X.M., Zha, Z. and Wu, F. (2020) Real-World Image Denoising with Deep Boosting. IEEE Transactions on Pattern Analysis and Machine Intelligence, 42, 3071-3087. [Google Scholar] [CrossRef] [PubMed]

为你推荐