针对传统高光谱影像低秩表示去噪方法无法保持影像多元几何结构信息的问题,提出一种基于局部超图拉普拉斯约束的高光谱影像低秩表示去噪方法。在低秩表示模型中增加超图拉普拉斯正则项,保持数据间多元几何流形结构;并对低秩模型系数矩阵增加稀疏和非负约束条件,进一步提高模型对影像局部信息的保持能力,使得模型不仅能够恢复具有低秩性质的影像信号分量,而且可以很好地保持影像的多元几何流形结构。在AVIRIS影像和ProSpecTIR-VS影像上的对比实验表明,所提方法更好地保持了影像的空间和光谱信息,有效地改善了高光谱影像去噪效果。 Aiming at the problem that traditional denoising method of hyperspectral image can not keep the multivariate geometric information of the image, a low-rank representation denoising method based on Laplacian of local hypergraphs is proposed. The hypergraph Laplacian regularity is added to the low rank representation model to maintain the multivariate geometric manifold structure among the data. The sparse and nonnegative constraints on the low rank coefficient matrix are added to further improve the model’s ability to retain the local information of the image So that the model not only can recover image signal components with low rank but also can maintain the multivariate geometry manifold structure of the image well. The comparison experiment between AVIRIS image and ProSpecTIR-VS image shows that the proposed method can better preserve the spatial and spectral information of the image and effectively improve the de-noising effect of hyperspectral image.