Vision-Based Fabric Defect Detection for Knitting
topic
Optical fabric inspection systems on circular knitting machines use line scan cameras with LED illumination to continuously image the knitted fabric surface at production speed, applying real-time image processing algorithms to detect holes, dropped stitches, tight courses, needle lines, and colour defects with defect position recorded relative to fabric roll length for quality mapping and loom stop triggering when defect frequency exceeds defined limits.
Role
Provides continuous automated fabric quality surveillance that catches defects missed by operator visual monitoring on high-speed machines with large fabric surface production rates, enabling earlier defect detection that reduces the length of defective fabric per fault occurrence and providing quantitative defect frequency data for quality management analysis and needle maintenance scheduling.