How Machines Automatically Spot Flaws in Fabric During Manufacturing

This patent describes an automated system for finding defects in fabric by cleaning it, taking pictures based on how a roller spins, and then using a computer to identify flaws.

What it covers

This patent outlines a method and system for automatically finding defects in fabric as it moves through a cloth inspection machine. First, a "noise removing device" (like a blower or cleaner, as per claim 6) prepares the fabric region to eliminate dust or debris that could look like a flaw (claim 1). Next, a camera takes pictures of the fabric, but it's triggered precisely by the rotation information of the machine's roller, ensuring consistent image capture (claim 1). Finally, a computer automatically analyzes these images to identify any defects (claim 1). If a defect is found, the system can slow down the fabric and mark the defect's exact location, even calculating its global position on the entire fabric roll (claims 2-4). For example, if a machine detects a tear, it can immediately slow down the fabric and spray a marker onto the edge of the fabric right next to the tear.

What it doesn't cover

• —Does not cover defect detection systems that do not include a dedicated noise removing device to preprocess the fabric.
• —Does not cover camera triggering for image acquisition that is not based on the rotation information of a cloth inspection machine's roller.
• —Does not cover manual human inspection of fabric defects, as the system is specifically designed for "automatically identifying" them.
• —Does not cover marking defects by translating the marking device, as it specifies marking the fabric edge "without translating the defect marking device."
• —Does not cover defect detection using non-visual sensors, as the core method relies on capturing and analyzing "one or more images."

The clever bit

The novelty lies in the precise coordination of steps: integrating a physical noise removal step, triggering camera image capture based on the exact rotation of the fabric roller, and then automatically identifying and pinpointing defects, even across multiple images through stitching.

Why it matters

Ensuring fabric quality is crucial in textile manufacturing, as defects can lead to wasted materials and costly returns. This type of automated system aims to improve efficiency and accuracy in quality control compared to traditional manual inspection. By precisely identifying and marking defects, manufacturers can streamline their processes, reduce human error, and ensure a higher standard of product before fabrics are cut and sewn.

Real-world examples

• 1.Automated textile inspection systems in fabric mills
• 2.Quality control lines for garment manufacturers
• 3.Smart factory solutions for textile production
• 4.Industrial fabric defect scanning machines