EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

MDHANet: Rethinking HOG as a Prior in Dense Networks with Self-Attention for Hyperspectral Image Classification

Hongwei Zhang iD, Yuanyuan Gui iD, Junjie Mou, Chao Lian iD

Sensors · 25.07.2026 · 26 (15) · 4731

DOI: 10.3390/s26154731 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, online erschienen am 25.07.2026. Verfasst von Hongwei Zhang und 3 weiteren Autor:innen. Beteiligte Einrichtung: School of Artificial Intelligence, China University of Mining and Technology-Beijing, Beijing 100083, China. Die Arbeit nennt 50 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Convolutional neural network (CNN)-based methods have been widely adopted in hyperspectral image (HSI) classification tasks and have demonstrated superior performance due to their potent nonlinear fitting capabilities. However, as network depth increases, existing methods suffer from significant issues such as gradient vanishing and information loss, which limit further improvements in model performance. Meanwhile, current approaches primarily design entirely new network architectures to extract abstract features but fail to fully exploit the explicit gradient prior information inherent in the data, which affects model performance in complex scenarios. To address these issues, this paper proposes a Multiscale Dense HOG-driven Self-Attention Network (MDHANet) that incorporates the Histogram of Oriented Gradients (HOG) as prior knowledge. First, the model employs three densely connected branches at different scales to obtain diverse multi-level feature representations through multi-scale feature reuse. Furthermore, since HOG can explicitly encode the distribution of gradient orientations and magnitudes in local neighborhoods and accurately characterize the feature distribution, a dynamic HOG-driven self-attention module is designed. Within this module, a HOG prior-based feature rearrangement strategy together with a dual-branch self-attention mechanism enables the network to more effectively learn the spatial information of the data, enhancing the extraction of discriminative spatial–structure features and further improving classification performance. Extensive experimental results on multiple datasets demonstrate that the proposed method outperforms existing state-of-the-art approaches. Extensive experimental results on four HSI datasets (IP, LK, HH, and HC) demonstrate that the proposed method outperforms existing state-of-the-art approaches, achieving OA of 99.13%, 97.32%, 99.02%, and 99.24%, respectively, with improvements of 0.83%, 0.41%, 2.69%, and 2.47% over the best competing methods.

Schlagwörter

Stichwörter der Autor:innen

dense connectionshistogram of oriented gradients (HOG)hyperspectral image classificationmultiscale feature extractionself-attentionspatial structure prior

Autor:innen und Institutionen

  • School of Artificial Intelligence, China University of Mining and Technology-Beijing, Beijing 100083, China.
  • State Key Laboratory of Space Information System and Integrated Application, Beijing 100095, China.
  • School of Information Science and Engineering, Northeastern University, Shenyang 110819, China.

ORCID-Profile: Hongwei Zhang, Yuanyuan Gui, Chao Lian

Zitieren

APA

Zhang, H., Gui, Y., Mou, J., & Lian, C. (2026). MDHANet: Rethinking HOG as a Prior in Dense Networks with Self-Attention for Hyperspectral Image Classification. Sensors, 26(15), 4731. https://doi.org/10.3390/s26154731

Vancouver

Zhang H, Gui Y, Mou J, Lian C. MDHANet: Rethinking HOG as a Prior in Dense Networks with Self-Attention for Hyperspectral Image Classification. Sensors. 2026;26(15):4731. doi:10.3390/s26154731

BibTeX
@article{Zhang2026,
  author = {Zhang, Hongwei and Gui, Yuanyuan and Mou, Junjie and Lian, Chao},
  title = {MDHANet: Rethinking HOG as a Prior in Dense Networks with Self-Attention for Hyperspectral Image Classification},
  journal = {Sensors},
  year = {2026},
  volume = {26},
  number = {15},
  pages = {4731},
  doi = {10.3390/s26154731}
}
Kostenlos starten →