Quiescent: A Wavelet De-Noising System

Period of Performance: 03/15/1995 - 09/15/1995


Phase 1 SBIR

Recipient Firm

THE Athena Group, Inc.
Principal Investigator


The removal of noise from complicated signals and images is a fundamentally important problem. Traditional methods, based on classical adaptive communication theory, have limited ability when presented with complex signals and images. Wavelets have recently been proposed as a potentially useful de-noising technology. Athena proposes QUIESCENT, a comprehensive wavelet-based de-noising study leading to a technology definition. Our approach is based on Athena's considerable experience in applying wavelets to medical images and wavelet processor designs. A highly qualified and experienced study team for QUIESCENT will employ advanced wavelet techniques and introduce new methodologies in multiscale representation on interval, interpolation, packets, local and best bases, to produce a powerful experimental and algorithm development workstation-based environment we call WES( Wavelet Experimental System). Within this environment, Athena will develop new soft and nonlinear thresholding, edge, and packet de-noising techniques, and experimentally study them over a large class of signals and images. Finally Athena will provide a definitive hardware implementation study which will lead to the definition of a real-time QUIESCENT processor.