The New Jersey Institute of Technology's
Electronic Theses & Dissertations Project

Title:	Design and implementation of generalized adaptive neural filters
Author:	Tam, Chun Yip
View Online:	njit-etd1998-108 (x, 54 pages ~ 1.7 MB pdf)
Department:	Department of Electrical and Computer Engineering
Degree:	Master of Science
Program:	Computer Engineering
Document Type:	Thesis
Advisory Committee:	Hou, Edwin (Committee chair) Ansari, Nirwan (Committee member) Zhou, MengChu (Committee member)
Date:	1998-01
Keywords:	Image processing. Adaptive filters. Computers--Design.
Availability:	Unrestricted
Abstract:	Generalized Adaptive Neural Filters (GANF) are a class of adaptive non-linear filters. This thesis presents a hardware implementation of GANF. Two designs are considered: the single neuron implementation and the multi-neuron implementation. The GANF design includes the window generator, threshold decomposer, training and filtering unit. The designs are verified through a logic design/simulation tool, Logic Works.