Semi-supervised Laplace Discriminant Embedding for Hyperspectral Image Classification

In order to extract effectively the discriminant characteristics of hyperspectral remote sensing image data, this paper presents a Semi-Supervised Laplace Discriminant Embedding (SSLDE) algorithm based on the discriminant information of labeled samples and the local structural information of unlabeled samples. The proposed algorithm makes use of the class information of labeled samples to maintain the separability of sample set, and discovers the local manifold structure in sample set by constructing Laplace matrix of labeled and unlabeled samples, which can achieve semi-supervised manifold discriminant. The experimental results on KSC and Urban database show that the algorithm has higher classification accuracy and can effectively extract the information of discriminant characteristics. In the overall classification accuracy, this algorithm is improved by 6.3%~7.4% compared with Semi-Supervised Maximum Margin Criterion (SSMMC) algorithm and increased by 1.6%~4.4% compared with Semi-Supervised Sub-Manifold Preserving Embedding (SSSMPE) algorithm.