Guangming Zhang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Guangming Zhang has a total of 3 co-authors. Here are the top co-authors by number of shared publications:
Juanjuan Ji
School of Electronic Information and Integrated Circuits, Anqing Normal University, Anqing, P. R. China
1 shared publication- Liangyou Lu
School of Information Science and Engineering, Southeast University, Nanjing 210096, China
1 shared publication - Peng Chen
Southeast University
1 shared publication
Latest Publications
- 27 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Geoscience and Remote Sensing Letters
- Vol. 23 (1545598X) , 1-5
Spatial–Spectral Texture Preserved Graph Total Variation Regularizer for Hyperspectral Image Denoising
Graph total variation (GTV) has been widely adopted for hyperspectral image (HSI) denoising. However, the fixed graph weights make it difficult to preserve or enhance important details. Moreover, such methods often focus on preserving spatial textures while neglecting the extraction of spectral texture information, failing to achieve a desirable balance between noise removal and spatial-spectral texture preservation. To address these shortcomings, this letter proposes a novel regularization, namely Spatial-spectral Texture Preserved GTV (SSTPGTV), which features two notable highlights. Firstly, we encode the spatial correlation between pixels via GTV in a band-by-band manner, and the graph weights are adaptively tuned according to the updated HSI data, ensuring that the refined details are preserved well. Secondly, we devise a novel weight learning strategy based on local line binary pattern (LLBP) to capture texture features along the spectral dimension. In addition, due to the fewer bands of the reduced image, SSTPGTV not only fully characterizes spectral correlation, but also maintains low computational complexity. Experimental results demonstrate that our method outperforms most mainstream denoising methods in terms of both performance and computational complexity.
- 27 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Geoscience and Remote Sensing Letters
- Vol. 23 (1545598X) , 1-5
Spatial–Spectral Texture Preserved Graph Total Variation Regularizer for Hyperspectral Image Denoising
Graph total variation (GTV) has been widely adopted for hyperspectral image (HSI) denoising. However, the fixed graph weights make it difficult to preserve or enhance important details. Moreover, such methods often focus on preserving spatial textures while neglecting the extraction of spectral texture information, failing to achieve a desirable balance between noise removal and spatial-spectral texture preservation. To address these shortcomings, this letter proposes a novel regularization, namely Spatial-spectral Texture Preserved GTV (SSTPGTV), which features two notable highlights. Firstly, we encode the spatial correlation between pixels via GTV in a band-by-band manner, and the graph weights are adaptively tuned according to the updated HSI data, ensuring that the refined details are preserved well. Secondly, we devise a novel weight learning strategy based on local line binary pattern (LLBP) to capture texture features along the spectral dimension. In addition, due to the fewer bands of the reduced image, SSTPGTV not only fully characterizes spectral correlation, but also maintains low computational complexity. Experimental results demonstrate that our method outperforms most mainstream denoising methods in terms of both performance and computational complexity.