Journal of Applied and Numerical Optimization

Full Text: PDF
DOI: 10.23952/jano.6.2024.3.02
Volume 6, Issue 3, 1 December 2024, Pages 323-338

 

Abstract. Linear hyperspectral unmixing (LHU) aims to extract endmembers from mixed pixels in hyperspectral images, which is also a large-scale problem as the number of pixels in hyperspectral images is very large. In this paper, by virtue of the Moreau envelope, we formulate the LHU as an unconstrained optimization problem. Then we adopt the proximal gradient descent method to solve the model, and consider a stochastic version of the method which is for dealing with large-scale scenario. Numerical results are provided to demonstrate the simplicity and efficiency of our proposed model.

 

How to Cite this Article:
X. Li, F. Xu, Y.-H. Dai, A stochastic proximal gradient method for linear hyperspectral unmixing, J. Appl. Numer. Optim. 6 (2024), 323-338.