Date:     Apr. 1, 3:30 pm in 4610 Engineering Hall

Title:    Linear Phase Perfect Reconstruction Filter Banks
          and Application in Image Compression

Speaker:  Trac Tran, ECE Department

Abstract:

There has been a tremendous growth in the field of filter banks (FB) and
multirate systems in the last fifteen years. These systems provide new and
effective ways to represent digital signals for computation, understanding,
and compression purposes. Filter banks find applications in virtually every
signal processing field. Obviously, of extreme importance is the ability to
design a FB that can fully exploit the properties and nature of a particular
signal or application.
 
This presentation focuses on the theory, structure, design, and implementation
of discrete-time FIR linear phase perfect reconstruction filter banks (LPPRFB)
with arbitrary M channels. All of the novel FB to be presented are obtained
from the attractive approach of FB's parameterization by lattice structures
based on the factorization of the analysis and synthesis polyphase matrices.
The lattice structurally enforces the LP and PR properties on the FB such that
these crucial properties are retained regardless of coefficient quantization.
 
M-channel LPPRFB can also be viewed as lembedded coder, can reach up to an
astounding 2.6 dB.  Future research directions and applications, including
video compression, analysis/compression/processing of medical signals, M-band
wavelets, and multicarrier modulation, will also be discussed.