FDICS - A VISION-BASED SYSTEM FOR THE IDENTIFICATION AND CLASSIFICATION OF FABRIC DEFECTS

In today's global market, the key to a manufacturing enterprise's succ ess lies in being competitive. To achieve this, the enterprise must ma ke use of state-of-the-art techniques such as computer-integrated manu facturing (CIM), just-in-time (JIT), and total quality management (TQM ). Computer vision is a relatively new technology that combines comput ers and video cameras to acquire, analyze, and interpret images in a w ay that parallels human vision. The objective of the research reported in this paper was to develop an automated defect-inspection and class ification system using the principles of machine vision, image-process ing, and pattern recognition. In this paper, the design, development, and use of a Fabric Defect Identification and Classification System (F DICS), a vision-based system for the identification and classification of fabric defects, is discussed. FDICS is made up of an image-acquisi tion module, a feature-extraction module, and a classification module. The image-acquisition module obtains the digitized image of the fabri c sample by using a video camera and stores it as an image file. The f eature-extraction module extracts the tonal and texture features from the image. The classification module classifies an unknown fabric samp le into one of five fabric classes based on a Mahalanobis classifier. FDICS has been shown to provide a higher percentage of correct classif ication than a similar system reported in the literature for defects c onsidered by both systems. The relative accuracy of using only either the tonal or the texture features was studied. The latter set of featu res gave a higher percentage of correct classification than the former ; however, the percentage was highest when both sets of features were used together.