Use pca in hyperspectral images with the toolbox Laurens van der Maaten in matlab

1

Someone could help me understand how I can use the pca of Laurens van der Maaten in matlab. The toolbox is in link I have the toolbox installed in matlab I have the hyperspectral image indian Pines leida

but how to use PCA for me to be able to identify the bands that do not serve me as those that have a lot of noise and are useless? in the help says

coeff = pca (X)

After reading the hyperspectral image, the dimensions are 145x145x200

Where it contains 200 bands but how can I use PCA for me to be able to identify the main components and the other rest of bands like the noisy ones using the graphic to visualize each band

I need to use figure (graph with imshow or imagesc) to visualize each band and see which ones are very noisy and then remove them

    
asked by semidf des 26.02.2017 в 21:51
source

0 answers