Peter Becker (Master thesis, Electrical and Computer Engineering, Drexel University, 2001)
A High Speed VLSI Architecture for the Discrete Haar Wavelet Transform

An alternative to strictly time- or frequency-domain analysis, wavelets provide a multiresolution analysis of signals for input to processing such as edge detection, compression, and pattern recognition. High speed VLSI architectures are necessary for real-time processing of large data sets. Presented here is such an architecture for the Haar wavelet transform, the simplest and easiest wavelet to understand. From this base model, a generalization of the architecture to other wavelets is discussed, as well.

IP cores for FPGA/ASIC, Wavelet Transform