Fuzzy Crystal Neural Network Implementation of Quantization Fuzzy Trigonometric Functions

摘要

The present invention relates to a method for clearing image quality by removing noise from an image, and in particular, a new form for processing an image signal by modifying an existing DBNN (Decisin Based Neural Network) model using a neural network and a fuzzy algorithm. To implement the FDNN (Fuzzy Decision Neural Network).;In other words, FDNN (Fuzzy Decision Neural Network), a new method of processing video signals, uses trigonometric quantization to fuzzy the pixels of the input image and replaces DWW (Weighted Fuzzy Mean) filter used in DBNN. Dynamic Weight Warping) algorithm was used.