reconstructing image that has been decompposed using svd

i have decomposed my image using svd... and modified the singular values..now i have three vectors..U S and V..can u pls help me to reconstruct my image.is there any keywords that would perform the inverse work of svd??

Answer 1

[Matt Tearle]

What do you mean by "the inverse of SVD"? The SVD is a decomposition: A = U*S*V' so reconstructing A just requires performing the multiplication.

[buddy]

thank you... i just thought there wil be some predefined keyword.. thanks it works in this way..

Answer 2

[Liz Hahn]

So A will be my image in principal component axis?

Answer 3

[Liz Hahn]

If I set data as my input image %svd: [u,s,v] = svd(data) data_new1 = u*s*v'; %pca [coef,score] = pca(data); data_new2 = (score')*data;

Are data_new1 = data_new2?