Developing a Neural Network to Act as a Noise Filter
Abstract
This study uses the neural network simulator called NETS to determine if neural networks could perform a non-linear filtering operation to remove noise from two-dimensional (2-D) data and produce a noise-free image. Application is geared toward the development and performance of neural network filters, including the development of an optional neural network architecture and the use of-criteria in determining how accurate the net filtered noise-to produce a noise-free image.
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 02, 1992
- Accession Number
- ADA259515
Entities
People
- Ramon Rivero