Application of Machine Vision in Textile Industry

The textile industry is an important representative of the industrial revolution, from handicraft workshops to factories stuffed with machines.

Today, the textile industry is deeply branded as a “traditional industry”.

Although the industry is traditional, but does not mean that it is not important: human life, clothing, food, housing and transportation.

The first, clothing, and textile is closely related.

Textile industry is not developed in the era, people wear simple, to meet the most basic warmth, shame function can be.

In modern society, the textile industry is greatly developed, the function of clothing, has not only been used to meet the most basic needs—- shame and warmth.

People pursue and enjoy beauty and individuality, and clothes, carrying people’s pursuit of beauty and individuality.

In order to meet the human pursuit of beauty and personality, as well as other areas of application, humans have woven a variety of textile fabrics.

Do you know how many kinds of textile fabrics have been created by human beings?

According to incomplete statistics, close to half a million kinds!

And each fabric, from a thread, to the fabric with color, need to be tested, the quality to meet the requirements, the consumer to buy.

The traditional approach, relying on manual, naked eye to check the line, the fabric on the defects.

Based on manual inspection, the cost is high and qualified textile inspection workers are not easily recruited. At the same time, the human fatigue state, prone to error, making defective products into the market.

Machine vision technology, applied to the textile industry in the field of inspection, has the characteristics of fast, real-time, accurate and efficient.

Machine vision inspection equipment replaces the human eye to complete surveillance, measurement and judgment, and has many application advantages such as non-contact, repeatable, reliable, high accuracy, continuity, high efficiency, and good flexibility.

It can realize the detection of defects in textiles such as broken warp, broken weft, rough knot, rough warp, weft file, loose edge, pilling, stain, hole, etc. It integrates various functions such as automatic detection, defect location, defect classification, fabric edge marking, quality assessment, etc., and can detect for many different textiles.

A typical machine vision fabric inspection system, consisting of mechanical, electronic, optical, computer and software components, is shown in the schematic diagram below.

In this system, the camera captures fabric image information in real time and transmits it to the backend system. After deep learning, the system is able to know whether the fabric in the current image has a defect, and if so, it sends a signal to make a corresponding response operation for that defect.

As a large textile country, China has tens of millions of various textile equipments in the country, and most of them, can be inspected by machine vision technology instead of manual inspection. The combination of machine vision technology and deep learning will allow the traditional industry of textile, with the wings of vision, to fly to a smarter and more efficient future.