An On-Line Monitoring System to Recognize Stitch Defects Through Thread Motion and High Speed Image Analyses

Abstract

A sensor system that inspects a garment as it is being sewn has the potential to eliminate much of the cost associated with inspection and to improve product quality. An economically feasible on-line stitch quality monitoring system is under development that utilizes commercially-available, low-cost piezoelectric sensors and a PC-based Linux data acquisition system. The sensors respond to the vibration caused by the thread motion and output a corresponding waveform used to study and model the formation of single stitches. As a result, the presence of periodic occurrences can be identified and attributed to proper stitch formation. Conversely, the absence of such events can be utilized to signal the presence of single stitch defects and diagnose their causes. Finally, high speed image analysis of the sewing threads has verified conclusions drawn from the output signals of the piezoelectric sensors and contributed to a better understanding of the dynamics involved in high speed sewing.