Directional selection using 2D complex discrete wavelet transform and its application for image processing

In this study, in order to realize the direction selectivity and shift invariability in all analysis levels simultaneously, we propose a method named the 2-dimensional complex discrete wavelet transform (CDWT) by using the RI-spline wavelet. We applied this method in fingerprint analysis. The results obtained can be summarized as follows: (1) The direction selectivity was realized by calculating the sum or difference between wavelet; coefficients that were obtained by the 2D CDWT. (2) 2D CDWT by RI-spline wavelet has been applied to the analysis of a fingerprint, and the direction of the 2D edges in all analysis levels could be extracted alternatively.