Real-Time White Noise Removal Using an Optimized Lightweight CNN Autoencoder for Embedded Image Processing

  • Unique Paper ID: 211128
  • Volume: 13
  • Issue: 6
  • PageNo: 3562-3567
  • Abstract:
  • White noise significantly impacts the performance of embedded vision systems employed in robotics, surveillance, healthcare devices, and IoT imaging modules. Conventional denoising methods, such as Gaussian, median, and wavelet filters, are computationally efficient but often result in the loss of important structural information. Deep-learning models provide superior reconstruction quality, yet their high computational requirements make them unsuitable for low-power embedded hardware. This paper presents a study and design of a lightweight Convolutional Neural Network (CNN) autoencoder specially optimized for real-time white noise removal on embedded platforms. The model integrates depthwise separable convolutions, skip connections, and 8-bit quantization to reduce memory consumption while maintaining strong denoising performance. Experimental evaluations indicate that compact CNN autoencoders can deliver considerable improvements in PSNR and SSIM while sustaining real-time operation, demonstrating their effectiveness for edge AI applications.

Copyright & License

Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

BibTeX

@article{211128,
        author = {BHOODARAPU RAJENDAR},
        title = {Real-Time White Noise Removal Using an Optimized Lightweight CNN Autoencoder for Embedded Image Processing},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {13},
        number = {6},
        pages = {3562-3567},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=211128},
        abstract = {White noise significantly impacts the performance of embedded vision systems employed in robotics, surveillance, healthcare devices, and IoT imaging modules. Conventional denoising methods, such as Gaussian, median, and wavelet filters, are computationally efficient but often result in the loss of important structural information. Deep-learning models provide superior reconstruction quality, yet their high computational requirements make them unsuitable for low-power embedded hardware. This paper presents a study and design of a lightweight Convolutional Neural Network (CNN) autoencoder specially optimized for real-time white noise removal on embedded platforms. The model integrates depthwise separable convolutions, skip connections, and 8-bit quantization to reduce memory consumption while maintaining strong denoising performance. Experimental evaluations indicate that compact CNN autoencoders can deliver considerable improvements in PSNR and SSIM while sustaining real-time operation, demonstrating their effectiveness for edge AI applications.},
        keywords = {White Noise, Embedded Vision, Lightweight CNN, Autoencoder, Image Denoising, Real-Time Processing},
        month = {November},
        }

Cite This Article

RAJENDAR, B. (2026). Real-Time White Noise Removal Using an Optimized Lightweight CNN Autoencoder for Embedded Image Processing. International Journal of Innovative Research in Technology (IJIRT), 13(6), 3562–3567.

Related Articles