Can we find the eigen vectors of an image which was generated through unknown orthogonal rotation? Let me explain my problem more clearly. I have one image which I got after applying Procrustes rotation. Then I tried to use 'eig' to find out its eigen vectors but unfortunately it said that matrix should be square. I need to find out the eigen vectors or principal component of that new image whose all axis are already orthogonal. Can anyone suggest me how to do it?