Real-time cloth defect detection method and system based on deep learning

Abstract

The embodiment of the invention discloses a real-time cloth defect detection method and system based on deep learning, and the method comprises the steps: 1, collecting different types of cloth defect images, and constructing a defect data set; 2, performing data expansion firstly, and then performing data expansion by means of a generative adversarial network; 3, carrying out labeling processing on the expanded defect data set; 4, constructing a deep learning target detection network to perform cloth defect detection; 5, training a cloth defect detection network; and 6, capturing images of the cloth in real time by using a camera, inputting the captured images into the trained cloth defect detection network, judging whether defects exist in the images, determining the types of the defects, positioning the defects, and finally storing a result into an output file. According to the method, manual design of features can be omitted, the robustness of a defect detection system is improved, the detection performance is greatly improved, manpower can be liberated, and the intelligent degree of the textile industry is further improved.

Description

technical field [0001] The invention relates to the technical field of cloth defect detection, in particular to a deep learning-based real-time cloth defect detection method and system. Background technique [0002] In cloth production, the quality of cloth is the most important thing, not only closely related to people's life, but also directly affects the development of the industry and the life of the enterprise. At present, in textile and garment production enterprises, professional cloth inspectors stand in front of the cloth inspection equipment, find out the defects of the cloth surface with the naked eye, and then mark or record the defects. However, there are many problems in relying on manual cloth inspection - low defect detection rate, difficult recruitment and high personnel costs, and the attention of cloth inspectors cannot be maintained continuously, and the cloth inspection speed is slow. If the cloth inspection time is too long, the cloth inspection workers...

AI Technical Summary

Problems solved by technology

However, there are many problems in relying on manual cloth inspection - low defect detection rate, difficult recruitment and high personnel costs, and the attention of cloth inspectors cannot be maintained continuously, and the cloth inspection speed is slow. If the cloth inspection time is too long, the cloth inspection labor will have problems Fatigue, the detection efficiency will be greatly reduced, very unstable

Therefore, it is a general trend to use computers to help humans detect defects. People take pictures of cloth and input them into the computer, and let the computer perform defect detection according to some image processing algorithms. However, traditional image processing generally requires manual processing for different scenarios. Design features, so the robustness is not enough, often changing a scene, the effect will be worse, the detection speed is not ideal, and there are often many types of cloth defects, it is difficult to manually design defect features

[0003] Relying on deep learning for defect detection can achieve better detection results, but it requires a large number of data samples as the basis. If the amount of data is not enough, the effect is often not very ideal. In actual production, there are often fewer data samples with defects. Doesn't work well with deep models

Images