For signal analysis and processing, a Fourier transform (FT) represents a signal in terms of its frequency contents. After the limitations of simple Fourier transformation were observed, especially for transient signals, windowed or short-time Fourier transforms and later wavelet transform (WT' s) were introduced~(1,2). A WT represents the signal in terms of a family of functions that is derived from a single basic function called a wavelet, by dilation (scaling) and translation (shift) operations~3. Fourier transformation gives global frequency information about the signal, whereas wavelet transformation gives local frquency information in the signal. With a WT, a scene (signal) can be decomposed into space-frequency (time-frequency)components. This space-frequency (time-frequency) representation nature of a WT leads to its many promising applications in multiresolution image analysis, data compression, pattern recognition, fractal analysis , transient signal, and image processing, etc.~(4-7)
展开▼