EUVIMEDEUVIMED entdecken
Crossref

Hybrid CNN-Autoencoder and Lempel-Ziv Approach for Efficient Lossless Image Compression

Esther J., Pathare Asha Gulab, Rani J.

i-manager’s Journal on Image Processing · 15.05.2026 · 13 (1) · 1

DOI: 10.26634/jip.13.1.1207 ↗

Kurz gefasst

Fachartikel in der Zeitschrift i-manager’s Journal on Image Processing, online erschienen am 05.05.2026. Verfasst von Esther J. und 2 weiteren Autor:innen. Die Arbeit nennt 12 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Image compression is the process of reducing the size of an image file without significantly compromising its quality. This is essential for saving storage space and enabling faster transmission across networks and devices. Lossless image compression techniques are particularly important in domains where preserving original image quality is critical, such as medical imaging, machine learning, and scientific research. Despite recent advancements, traditional compression methods face several challenges, including quality degradation, inefficient handling of complex image structures, limited support for metadata and transparency, high computational costs, and poor performance under extreme conditions. To address these limitations, this study proposes a hybrid approach that combines deep learning with classical compression techniques for efficient lossless image compression. The proposed method employs a Residual Convolutional Neural Network Autoencoder (RCNN-AE) for effective feature extraction and dimensionality reduction. Residual connections enhance the learning of complex image features while minimizing reconstruction loss. The compact feature representations generated by the RCNN-AE are further compressed using the Lempel–Ziv (LZ) algorithm, a well-established lossless compression technique. This dual-stage process significantly improves compression efficiency while ensuring complete preservation of image quality. The decompression module mirrors the compression pipeline, enabling accurate reconstruction of the original image from the compressed representation. Experimental evaluation on the Kodak dataset demonstrates that the proposed Hybrid RCNN-AE + LZ approach achieves superior compression ratios and high reconstruction fidelity compared to conventional methods, thereby validating its effectiveness for real-world lossless image compression applications.

Zitieren

APA

J., E., Gulab, P. A., & J., R. (2026). Hybrid CNN-Autoencoder and Lempel-Ziv Approach for Efficient Lossless Image Compression. i-manager’s Journal on Image Processing, 13(1), 1. https://doi.org/10.26634/jip.13.1.1207

Vancouver

J. E, Gulab PA, J. R. Hybrid CNN-Autoencoder and Lempel-Ziv Approach for Efficient Lossless Image Compression. i-manager’s Journal on Image Processing. 2026;13(1):1. doi:10.26634/jip.13.1.1207

BibTeX
@article{J2026,
  author = {J., Esther and Gulab, Pathare Asha and J., Rani},
  title = {Hybrid CNN-Autoencoder and Lempel-Ziv Approach for Efficient Lossless Image Compression},
  journal = {i-manager’s Journal on Image Processing},
  year = {2026},
  volume = {13},
  number = {1},
  pages = {1},
  doi = {10.26634/jip.13.1.1207}
}
Kostenlos starten →