Development of computer vision tools for detecting defects in garment manufacturing in outsourcing factories

• Guseva M.A.

  Marina A. Guseva.The Kosygin State University of Russia. Moscow. Russia

• Andreeva E.G.

  Elena G. Andreeva. The Kosygin State University of Russia. Moscow. Russia

• Yu. V. Rogozhina

  uliya V. Rogozhina. The Kosygin State University of Russia. Moscow. Russia

he global strategy for the development of industrial enterprises involves the creation of
the latest digital platforms that unite all participants in production cycles, including those involved in
the sale of finished products to consumers. The Russian apparel industry has historically developed
strong production links with outsourcing firms in Southeast Asia, of which China factories are a pri-
ority. The experience of sewing business projects in cooperation with outsourcing contractors has
shown that in the context of geopolitical constraints the most effective results are obtained by compa-
nies that use the advantages of the industrial Internet. During the pandemic, the difficulties of inter-
action between customers and contractors increased, both at the interstate level and within the con-
tractor countries. At sewing enterprises, the stage of technological control turned out to be the most
unsuitable for uninterrupted functioning under conditions of restrictions. The classic approach to
evaluating the level of clothing manufacturing is based on contact measurements of semi-finished
products and finished products. Until now, the industry lacked experience in remote inspection of
garment manufacturing quality using computer vision. The article presents innovative developments
of the authors from The Kosygin Russian State University to optimize the quality control process of
garment production using computer vision systems at the stage of final inspection. The proposed
method has been tested at outsourcing enterprises in China.
Keywords: scanning, machine vision, quality of sewing products