Extraction of subimages by lifting wavelet filters

This paper proposes a method for extracting subimages from a huge reference image by learning lifting wavelet filters. Lifting wavelet filters are biorthogonal wavelet filters containing free parameters developed by Sweldens. Our method is to learn such free parameters using some training subimages so as to vanish their high frequency components in the y- and xdirections. The learnt wavelet filters have the feature of training subimages. Applying such wavelet filters to the reference image, we can detect the locations where the high frequency components are almost the same as those of the target subimage.