This is the authors' abstract. We don't add key points for this paper.
Hyperspectral image (HSI) classification is an important technique in the field of remote sensing. Recent years have seen the use of deep learning methods as powerful feature extractors for hyperspectral image classification. While convolutional neural networks have achieved high accuracy and efficiency, the problem of misclassifying adjacent ground object boundary pixels remains prominent due to noise and band correlation between pixels. Additionally, selecting information-rich spectral-spatial kernel features poses a challenge. To address these issues, this paper proposes a spatial kernel residual network with an attention module for feature extraction, as well as a denoising module to remove noise. The proposed method was evaluated on two public HSI datasets, demonstrating its effectiveness and excellent classification performance.
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Fan et al. (2024) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: