Translation Invariance and Sampling Theorem of Wavelet

Q Wang,LN Wu

DOI: https://doi.org/10.1109/78.839994

IF: 4.875

2000-01-01

IEEE Transactions on Signal Processing

Abstract:The sampling theorem for wavelet spaces built by Waiter lacks the translation invariance except for Waiter's weak translation invariant wavelet, i.e., Meger's wavelet. Indeed,we must know a priori the shift offset a in the samples {f(n + a);n is an element of ZZ}; otherwise, the waveform cannot be recovered since the Interpolation function is dependent on this offset. in this correspondence, we generalize our metric functional to metrize weak shiftability and find a somewhat superising result that the B spline wavelets of order n greater than or equal to 3 are degenerate shiftable. Thus, we can recover approximately the waveform by double sampling without ang information on shift offset a.