Wavelets facilitate the combination of frequency domain and spatial domain
representations of a signal. The technique allows the treatment of signals
which are unsuitable for traditional fourier signal analysis - generally
because such assumptions as continuity and stationarity are not required.
The name wavelet is meant to suggest the wave packet, which is well
localised in both spatial and frequency domains simultaneously. Wavelet
analysis was pioneered by
Daubechies (1992) (see mathematical text by e.g.
Meyer 1993). Further information on the our wavelet analysis is presented
elsewhere (Hurley et al. 1997).
Figure 1:
Trigger No. 1606: Original (solid line) and reconstructed (dashed
line) time history are almost indistinguishable. The peaks identified by the
algorithm are marked
