Abstract:
The crosscorrelation estimation methods for two-dimensional multiscale signals are considered. It is shown that the
wavelet decomposition allows one to estimate the correlation for each scale individually. The structure orientation for different scales can be taken into account when applying anisotropic wavelets. The efficiency and limitations of these methods are illustrated using several test examples.
Keywords:wavelet correlation, signal processing, crosscorrelation, wavelet decomposition, anisotropic wavelets.