Karhunen-Loeve Transform of a Set of Rotated Templates

IEEE Trans. on Image Processing, 2003
We propose a novel method for efficiently calculating the eigenvectors of uniformly rotated images of a set of templates. As we show, the images can be optimally approximated by a linear series of eigenvectors which can be calculated without actually decomposing the sample covariance matrix.


<a href="http://prints.vicos.si/publications/70">Karhunen-Loeve Transform of a Set of Rotated Templates</a>