Image processing based production flaw detection in knitting

Abstract

One of the major problems in production is that the number of faults occurred while producing a fabric, since it directly influences productivity. To overcome the losses, this paper provides a system with special surveillance to detect the knitting process, identify and locate faults during manufacturing, by inspecting the fabric. The system also gives the user a valuable set of data related with production. This examination has demonstrated that picture handling can possibly give solid estimations to recognizing abandons in sewed textures. On-line texture deformity location was tried consequently by contrasting texture pictures caught by a computerized camera. In this way, it is demonstrated that the created picture catching and investigation framework is equipped for halting the round sewing machine by utilizing processor when a deformity is caught by the web camera.