Title :
Kernel-Based Linear Spectral Mixture Analysis
Author :
Liu, Keng-Hao ; Wong, Englin ; Du, Eliza Yingzi ; Chen, Clayton Chi-Chang ; Chang, Chein-I
Author_Institution :
Dept. of Comput. Sci. & Electr. Eng. Dept., Univ. of Maryland, Baltimore, MD, USA
Abstract :
Linear spectral mixture analysis (LSMA) has been widely used in remote sensing community for spectral unmixing. This letter develops a promising technique, called kernel-based LSMA (KLSMA), which uses nonlinear kernels to resolve the issue of nonlinear separability arising in unmixing and further extends several commonly used LSMA techniques to their kernel-based counterparts. Interestingly, according to experiments conducted for real hyperspectral and multispectral images, KLSMA is more effective than LSMA when data samples are heavily mixed.
Keywords :
geophysical image processing; geophysical techniques; remote sensing; spectral analysis; LSMA techniques; hyperspectral image; kernel-based linear spectral mixture analysis; multispectral image; nonlinear kernel analysis; remote sensing; Hyperspectral imaging; Kernel; Principal component analysis; Support vector machines; Training; Fully constrained least squares (FCLS); LSMA; kernel-based linear spectral mixture analysis (LSMA) (KLSMA); least squares orthogonal subspace projection (OSP) (LSOSP); nonnegative constrained least squares (NCLS);
Journal_Title :
Geoscience and Remote Sensing Letters, IEEE
DOI :
10.1109/LGRS.2011.2162088
