Detection and Classification of Weft Knitted Fabrics Defects Using Gabor Wavelet

The globalization of competition, the complexity of the economy and the plethora of information available today place companies in a more than shifting context with which they must cope. Accelerating change becomes a constant feature of business life. Companies need methodological help to process a lot of information, In today's competitive world, customers are demanding better quality products with fast and reliable deliveries. To meet this demand, new manufacturing technologies are developing rapidly, resulting in new products and improvements in manufacturing processes. Today’s challenging world demands minimum loss and waste from industries. Moreover, it has to ensure the required quantity and quality with customer delivery lead time. A Circular weft knitting machine contains different parts such as needles, cams, sinkers, Fabric takedown mechanism, creel, a yarn metering and storage device, yarn breakage indicator, feeders and lubrication system. All those machine parts are responsible to increase or decrease the productivity of weft knit fabric production as well as the fabric quality. The Circular weft Knitting Machine has to stop when defects occurred and then faults are corrected, which results in a loss in time and efficiency in order to be ready to meet customer requirements; the goal is to quickly provide products that combine quality and competitive price. In this sense, effective monitoring is required to avoid defects and maintain high productivity and customer required quality. The purpose of this study is to identify and analyze weft knitted fabric defects on the weft circular knitting machine of knitting industries. This paper describes a computer vision-based fabric inspection system implemented on a weft circular knitting machine to detect defection and classification of the weft-knitted fabric defects under construction. We using Gabor Wavelets that have been successfully applied to various machine vision applications such as Texture segmentation, Edge detection, and Boundary detection, a multi-scale and multi orientation Gabor filter scheme simulates the human eye and that’s applied to the weft-knitted fabric under construction. On-line weft knitted fabric defect detection was tested automatically by analyzing fabric images captured by a digital camera using Gabor wavelets and classification (Identification) these fabric defects to known classes. We succeeded to detect weft knitted fabric defects and classify defect’s on the weft circular knitting machine at the same time the machine stops to correct the fabric defect to achieve customer required quantity and quality. As well as we cancel the fabric inspection process that means, saving money, time, manpower which leads to reducing production lead time and cost