A Semisupervised Spatial Spectral Regularized Manifold Local Scaling Cut With HGF for Dimensionality Reduction of Hyperspectral Images (original) (raw)

Hyperspectral images (HSI) present challenges in classification due to high dimensionality and limited labeled samples. This paper proposes a semi-supervised graph-based dimensionality reduction method, named 'semi-supervised spatial spectral regularized manifold local scaling cut' (S3RMLSC), which effectively utilizes both labeled and abundant unlabeled data by retaining the distribution of original data. The method consists of smoothing using a hierarchical guided filter, constructing linear patches from nonlinear manifolds, and optimizing a semi-supervised dissimilarity projection matrix. Validation is conducted using real-world HSI datasets to demonstrate the algorithm's effectiveness in addressing issues in HSI data analysis.