Non-negative matrix factorization method applied to hyperspectral image processing

Abstract

The present invention discloses a Non-negative Matrix Factorization (NMF) based on sparse and correlation constraints and the method is applied to processing of decomposition of mixed pixels of a hyper-spectral remote sensing image. According to the method, finally, a given non-negative matrix Vm*n is factorized into a product of a basis matrix Wm*r and a coefficient matrix Hr*n, i.e. Vm*n is approximately equal to Wm*r Hr*n ; firstly, the non-negative matrix V is selected, W and H are randomly initialized, then the minimum correlation constraint is applied to the coefficient matrix H in a target function, the sparse constraint is applied to the basis matrix W and the coefficient matrix H and then iteration is carried out according to an iteration formula until the matrices W and H are converged.

Description

technical field [0001] The invention relates to a hyperspectral image processing method, specifically a non-negative matrix factorization algorithm (Non-negative Matrix Factorization, NMF) based on sparsity and correlation constraints, and belongs to the technical field of hyperspectral remote sensing image processing. Background technique [0002] Remote sensing technology (Remote Sensing, RS) refers to the comprehensive observation technology of the earth's surface from long-distance space (aerospace) or outer space (aviation) that began in the 1960s. Remote sensing refers to indirect observation and monitoring from long-distance or outer space, without touching the target, to obtain relevant information about the target, phenomenon and area from the optical point of view, so as to carry out data fusion analysis and inference, and finally achieve the goal of obtaining the required target information. means, technology and science. Hyper-spectral image remote sensing techn...

AI Technical Summary

Problems solved by technology

However, the non-negative matrix factorization algorithm has the limitations of slow iterative convergence and easy to fall into a local optimal solution.

Images