Detection and Removal of Defects in Apparel Production

Abstract : Defects in fabric have been and continue to be a major source of seconds in finished garments. These, defects persist despite several visual inspections and intensive efforts to remove defective parts during, sewing operations. The increase use of automation in assembly steps will intensify the problem of detection and removal of fabric defects in cut-parts. A workstation utilizing machine vision has been designed and constructed to detect and remove defective cut-parts prior to initiation of the assembly operations. The workstation employs two vision systems--an area camera and a line camera--to inspect parts on a conveyor belt both statically and dynamically. The color of the parts is also determined and the area and perimeter are measured to detect improperly cut parts. The acceptable parts are then stacked in a manner suitable for input to an automated sewing station. The workstation should permit placing into the assembly operations a set of defect free, properly cut and color matched parts. It is estimated that this cut-part inspection system will reduce defects in finished garments by approximately 50% and should greatly simplify the labor intensive and costly fabric defect control systems currently in place in most apparel plants. The completed cut-part inspection workstation was demonstrated at the DLA Apparel Manufacturing Technology Center at Southern College of Technology.