ACC
Research on Image Denoising Method Based on Convolutional Neural Network
TOTAL VIEWS: 927
Abstract
With the continuous advancement of deep learning technology, image denoising has gradually evolved from traditional methods to complex deep model architectures. This paper systematically reviews the technical evolution of deep neural networks in image denoising, from residual learning, multi-scale fusion, and attention mechanisms to generative adversarial networks, analyzing the innovative features of various architectures and their contributions to improving naturalness, detail restoration, and robustness. It discusses future development directions such as model lightweighting, cross-modal fusion, and self-supervised learning, emphasizing the role of multi-domain, multi-task, and multi-scale fusion strategies in enhancing model generalization ability and deployment efficiency. Simultaneously, future research needs to focus on model interpretability, privacy protection, and industry adaptability to provide theoretical support for the widespread application of deep denoising technology. The continued innovation of deep denoising technology promises to achieve higher performance and stronger adaptability in various practical scenarios such as medical imaging, autonomous driving, and remote sensing monitoring, leading intelligent visual perception to a higher level.
Keywords
Deep learning; Image denoising; Multi-scale fusion; Attention mechanism; Generative adversarial networks; Self-supervised learning; Model lightweighting; Cross-modal fusion
References
[1] Li M. Research on image denoising method based on convolutional neural network [Master's thesis]. Shandong Normal University; 2024.
[2] Zhu Y, Lin G. Real image denoising algorithm based on convolutional neural network. J Chengdu Univ Technol. 2024;27(06):45-50,62.
[3] Jiang W, Bu Y. Image denoising network integrating CNN and Transformer. Appl Comput Syst. 2024;33(07):39-51.
[4] Wang S. Research on denoising algorithm of low-dose CT images based on convolutional neural network [Master's thesis]. North University of China; 2024.
[5] Tang Y. Image defogging processing based on region segmentation and multi-scale convolutional neural network [Master's thesis]. Yancheng Institute of Technology; 2024.
[6] He T, Wang C, Wu G. Research on denoising algorithm for low-light images based on convolutional neural network. Sens Microsyst. 2023;42(12):64-7.
[7] Ma Z, Tian B. A new method for SAR image denoising based on convolutional neural network. Fire Control Radar Technol. 2023;52(01):15-20.
[8] Wang J. Research on image real noise removal method based on convolutional neural network [Master's thesis]. Harbin Institute of Technology; 2022.
How to cite this paper
Research on Image Denoising Method Based on Convolutional Neural Network
How to cite this paper: Shiqi Wang. (2025) Research on Image Denoising Method Based on Convolutional Neural Network. Advances in Computer and Communication, 6(3), 102-106.
DOI: http://dx.doi.org/10.26855/acc.2025.07.001
More Articles
- Research on Enhancing ERP System Efficiency Through AI in Cross-border Supply Chain Environments
- Generative AI-driven Visual Advertising Creativity Generation and Style Consistency Modeling
- Generative AI-driven Visual Advertising Creativity Generation and Style Consistency Modeling
- Analysis of Implementation Pathways for Generative AI in Cross-language Code Migration and Compatibility Testing
- Research on Intelligent Marketing Strategy Optimization Based on Generative AI and User Behavior Prediction
- Research on the Application of Sentiment Analysis in Customer Segmentation and Precision Marketing