Fuzzy logic spots the defects in fabric

Indian researchers from the Kongu Engineering College (Perundurai, India) have developed an automated system to inspect the quality of knitted fabric produced by the textile industry.

To do so, a video of the moving fabric that is being produced is first captured by an array of cameras after which individual frames from the video are processed to improve the quality of the images.

Textural features extracted from them are then analyzed using a fuzzy logic system implemented with National Instruments (Austin, TX) LabView fuzzy system designer toolbox to identify any faults in them. After the analysis, the percentage of faults in the fabric is then determined.

The researchers claim that their proposed system is over 20 percent better at identifying faults in the fabric and over 10 percent better result at classifying them than existing microcontroller-based automated fabric inspection methods.

More information on the system can be found here.

-- Dave Wilson, Senior Editor, Vision Systems Design