EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

A Heterogeneous Modular Framework for Pre-Trained Image Dehazing Models Based on Haze Level Clustering

Cheng-Hsiung Hsieh, Xin-Rui Lin, Wei-Cheng Liao, Yung-Fa Huang

Electronics · 16.04.2026 · 15 (8) · 1676

DOI: 10.3390/electronics15081676 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Electronics, online erschienen am 16.04.2026. Verfasst von Cheng-Hsiung Hsieh und 3 weiteren Autor:innen. Die Arbeit nennt 37 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

While pre-trained deep learning models have significantly advanced image dehazing, their restoration performance often fluctuates substantially across varying haze densities, leading to inconsistent performance across diverse atmospheric conditions. To address this limitation, this study introduces a performance analysis approach based on Haze Image Clustering (HIC) to systematically evaluate the specialized strengths of various state-of-the-art models within specific haze-level intervals. Building upon these evaluations, we propose a heterogeneous modular framework equipped with a dynamic switching mechanism that adaptively activates the optimal pre-trained module for each detected haze level. Extensive experiments conducted on the OTS and ODF benchmark datasets demonstrate that while individual models exhibit regional performance drops, the proposed framework consistently maintains superior performance across all haze intensities. Quantitative results indicate that the proposed modular network achieves a significant PSNR improvement of up to 6.946 dB compared to DehazeFlow. Furthermore, regarding the no-reference Dehazing Quality Index (DHQI), our framework attains a top score of 68.448, surpassing the best individual baseline. These findings validate that the proposed strategy effectively enhances both restoration fidelity and visual naturalness without the need for additional training or fine-tuning, offering a robust and computationally efficient solution for real-world image dehazing.

Zitieren

APA

Hsieh, C., Lin, X., Liao, W., & Huang, Y. (2026). A Heterogeneous Modular Framework for Pre-Trained Image Dehazing Models Based on Haze Level Clustering. Electronics, 15(8), 1676. https://doi.org/10.3390/electronics15081676

Vancouver

Hsieh C, Lin X, Liao W, Huang Y. A Heterogeneous Modular Framework for Pre-Trained Image Dehazing Models Based on Haze Level Clustering. Electronics. 2026;15(8):1676. doi:10.3390/electronics15081676

BibTeX
@article{Hsieh2026,
  author = {Hsieh, Cheng-Hsiung and Lin, Xin-Rui and Liao, Wei-Cheng and Huang, Yung-Fa},
  title = {A Heterogeneous Modular Framework for Pre-Trained Image Dehazing Models Based on Haze Level Clustering},
  journal = {Electronics},
  year = {2026},
  volume = {15},
  number = {8},
  pages = {1676},
  doi = {10.3390/electronics15081676}
}
Kostenlos starten →