Public Release: 

On-Line Inspection System Uses Neural Nets, Wavelets & Fuzzy Logic To Improve Textiles

Georgia Institute of Technology

An automated on-line inspection system that uses advanced vision technology, neural networks, fuzzy logic and wavelets to quickly identify defects in fabric has been successfully tested at an Alabama textile manufacturing plant. The system could cut inspection costs and reduce the amount of off-quality production.

Developed by researchers at the Georgia Institute of Technology, the equipment could ultimately be the basis for an integrated electronic feedback system that would monitor and control quality processes throughout the manufacturing cycle. A West Virginia maker of textile equipment has licensed the technology with plans to produce a commercial system.

"Quality is extremely important to our industry," said Bart Krulic, sales representative for the Industrial Fabrics Group of Johnston Industries, which tested the system at its factory in Phenix City, Alabama. "We think the potential of this system is tremendous because it would be a low-cost measure that could be added to looms to improve quality."

Developed with support from the National Textile Center University Research Consortium, the system has operated on a loom at the company's Southern Phenix plant since July 1996. Krulic said the system has performed well, and he believes its use would give the company a strong competitive advantage by ensuring consistent error-free inspection.

"This is a good opportunity to take our quality processes to the next stage," he added. "You could reduce the labor and human error, and improve the quality. Improving quality and reducing costs will make a company more competitive in world markets."

Inspecting textile fabrics can be challenging because the industry uses many types of yarns and weave patterns, quality standards can differ by company, and existing systems rely on manual processes subject to human error. Inspection is expensive for the textile industry, costing as much as a million dollars per year at some manufacturing plants.

The value of fabric affected by defects can also be significant, since recurring problems in high-speed looms can damage thousands of yards of fabric if not quickly found and corrected. Most manufacturing plants employ inspectors to watch for off-quality problems during weaving, but some defects are still not identified until final inspection.

The new Georgia Tech system automatically identifies defects as the fabric comes off the loom, allowing the manufacturer to immediately correct process problems.

"This system would allow you to prevent the production of defects by correcting problems more quickly on the machine," explained Dr. Lew Dorrity, a professor in Georgia Tech's School of Textile and Fiber Engineering. "You would improve the quality of the manufacturing process, not just determine what quality exists in the finished products."

The Georgia Tech system uses a special lighting arrangement and a set of high-speed cameras to scan fabric as it winds onto a take-up roll after weaving. A computer analyzes the information provided by the vision system using sophisticated techniques that identify abnormal patterns and determine whether they should be considered defects.

Besides detecting off-quality fabric, the inspection system can also provide information that will help companies pinpoint the factors that cause defects. And the system could provide a detailed record of weaving quality that apparel manufacturers can use to optimize their use of fabric.

To make the system work reliably, the researchers combined cutting-edge software technology able to rapidly process information, learn from its mistakes and mimic human decision- making techniques.

"We have developed new and innovative software algorithms for the detection and classification of defects," explained Dr. George Vachtsevanos, a professor in Georgia Tech's School of Electrical and Computer Engineering. "We extract signatures from the images that are characteristic of the type of defect that might be present. We use a new wavelet/neural network construct for this signature extraction, along with fuzzy logic decision support systems. The software integrates learning and optimization tools that avoid false alarms and improve the recognition accuracy."

To operate in the demanding textile environment, the system has to withstand vibration, lint and other severe operating conditions. It also needed to operate reliably and be inexpensive enough to be widely adopted. Dorrity said the prototype Georgia Tech system, built from commercially-available lighting, vision and computer equipment, performed well in the manufacturing environment.

The system's technology has been licensed to Appalachian Electronic Instruments, Inc., a West Virginia manufacturer of textile-related equipment. The company expects to turn the prototype into a commercial system that can be retrofitted to existing looms and installed in new machines.

Part of that development will involve designing customized systems and reducing the overall cost of the system. Also needed are operator interface techniques that will allow plant personnel to operate the system and customize inspection parameters for different fabrics and weaves.

Current research and development activities at Georgia Tech, under continued sponsorship by the National Textile Center, are focusing on improving the software routines to detect a wider class of defects for a variety of fabric styles and the introduction of custom designed electronics to replace the personal computer platform.

Digital signal processors offer a cost-effective, fast parallel processing capability that promises to reduce defect recognition times by orders of magnitude. They open up a new window of opportunity for the utility of these innovative technologies to such critical quality inspection tasks as paper web monitoring, metal fabrication, glass production, and food processing. These industries are relying upon improved product quality and increased productivity in order to maintain their competitive advantage in a global economy

"We believe there are many generic features in these technologies that could be used in other industries that need this inspection capability," Vachtsevanos added. "There are possibilities far beyond the textile industry.

He said the project has moved from concept to beta testing in less than three years because of a close partnership between Georgia Tech, Johnston Industries and Appalachian Electronic Instruments, Inc.


430 Tenth St. N.W., Suite N-112
Georgia Institute of Technology
Atlanta, Georgia 30318

MEDIA RELATIONS CONTACTS: John Toon (404-894-6986)
or Amanda Crowell (404-894-6980); Internet:
FAX: (404-894-6983)

TECHNICAL CONTACTS: Dr. Lew Dorrity (404-894-9076);
Dr. George Vachtsevanos (404-894-6252);
Bart Krulic (334-768-1075); FAX: (334-768-1148).

VISUALS: Color slides showing researchers with loom and inspection equipment, regular plant workers with loom and inspection equiment.

WRITER: John Toon


Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.