Neural and bayesian networks based Fabric defect detection with the Microcontroller in manufacturing Of textiles

Abstract

In Electrical Engineering Field Image processing is any form of signal newlineprocessing for which the input is an image such as a photograph or video frame newlinethe output of image processing may be either an image or a set of newlinecharacteristics or parameters related to the image Many techniques have been newlineused in textile field for reducing the fault on the fabrics The fault on fabrics newlinesuch as holes scratch stains missing yarn knots gout has been reduced by newlineusing the image processing technique along with Neural Networks Bayesian newlineNetworks and Microcontroller process Noise removal or De noising is an newlineimportant task in image processing to recover a signal that has been corrupted newlineby noise newlineThis Thesis Focuses on the Fabric defects Some Fabric sample images newlineare taken from Textile Industries and are processed by software Then the newlineimages are processed by Thresholding and histogram process The fabric newlineoutput image is processed by neural network The term neural network was newlinetraditionally used to refer to a network or circuit of biological neurons The newlinemodern usage of the term often refers to artificial neural networks which are newlinecomposed of artificial neurons or nodes In this work the neural networks used newlineare Feed forward and Back propogation networks which are trained to identify newlinethe faults present on the fabrics The neural network output is compared with newlineBayesian Network newline