Real-Time Image Defect Detection System of Cloth Digital Printing Machine

Area of Science:

• Textile Manufacturing
• Computer Vision
• Image Processing

Background:

• Digital printing of fabrics can result in surface defects like white silk, spots, and wrinkles.
• Accurate detection of these defects is crucial for maintaining fabric quality and reducing production waste.

Purpose of the Study:

• To develop an automated system for detecting surface defects in digital printed fabrics.
• To improve the accuracy and efficiency of defect detection in industrial textile production.

Main Methods:

• Image registration using the speeded up robust features (SURF) algorithm.
• Employing a bidirectional unique matching method to minimize mismatch points for accurate image registration.
• Utilizing a difference algorithm to extract defect information from registered images.

Main Results:

• The developed system achieved a 98% detection accuracy for surface defects on printed fabrics.
• The improved SURF algorithm demonstrated a higher detection rate and faster detection speed compared to existing methods.
• Experimental validation using multiple images confirmed the system's performance.

Conclusions:

• The proposed surface defect detection system effectively identifies fabric flaws in digital printing.
• The enhanced SURF algorithm offers a robust and efficient solution for automated textile inspection.
• The system meets the requirements for practical industrial applications, enhancing quality control.