Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Cloth defect detection method and system

A defect detection and cloth technology, applied in the field of computer vision, to achieve the effect of convenient subsequent segmentation, high computing efficiency, and enhanced reconstruction effect

Pending Publication Date: 2022-07-08
HUAZHONG UNIV OF SCI & TECH
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] In view of the above defects or improvement needs of the prior art, the present invention provides a cloth defect detection method and system to solve the technical problem that the prior art cannot accurately detect cloth defects with high computing efficiency

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Cloth defect detection method and system
  • Cloth defect detection method and system
  • Cloth defect detection method and system

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0049] A cloth defect detection method, such as figure 1 shown, including the following steps:

[0050] S1. Input the cloth sample to be tested into the pre-trained lightweight generative adversarial network to obtain a reconstructed image of the cloth sample to be tested; the obtained reconstructed image is a defect-free image of the cloth sample to be tested.

[0051] Specifically, the training method for the above-mentioned lightweight generative adversarial network includes: duplicating the collected non-defective cloth sample set into two copies, and adding random noise to the samples in one of the non-defective cloth sample sets as a lightweight generative confrontation The input of the network, another non-defective cloth sample set does not do any processing, and is used as the output of the lightweight generative adversarial network, through the game between the light-weight generator and the discriminator, in order to fight against the light-weight generative adversa...

Embodiment approach

[0071] In another optional implementation manner, by comparing the structural similarity between the cloth sample to be tested and its reconstructed image, the cloth defect is detected according to the degree of similarity. If the difference between the cloth sample to be tested and its reconstructed image is less than the preset difference value, the cloth sample to be tested has no defects; otherwise, the cloth sample to be tested is defective, and the area on the cloth sample to be tested is different from its reconstructed image is the defect location.

[0072] In order to further illustrate the performance of the cloth defect detection method provided by the present invention, the convolutional denoising autoencoder model (CDAE), the multi-scale convolutional denoising autoencoder model (MSCDAE), the U-shaped convolutional denoising autoencoder The encoder model (UCDAE) and the lightweight generative adversarial network (LCD-GAN) proposed by the present invention are trai...

Embodiment 2

[0082] A cloth defect detection system, comprising:

[0083] The image reconstruction module is used to input the cloth sample to be tested into the pre-trained lightweight generative adversarial network to obtain the reconstruction map of the cloth sample to be tested;

[0084] The defect detection module is used to compare the difference between the cloth sample to be tested and its reconstructed image to detect cloth defects;

[0085] The above lightweight generative adversarial network includes a lightweight generator and a discriminator; the lightweight generator is used to generate a reconstructed image of the input cloth sample; the discriminator is used to determine whether the input image is a real cloth sample or a generated reconstructed image ;

[0086] Among them, the lightweight generator includes: encoder and decoder;

[0087] The encoder includes a convolutional layer and a bottleneck module connected in series; the bottleneck module includes a plurality of bot...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a cloth defect detection method and system, and belongs to the field of computer vision, and the method comprises the steps: S1, inputting a to-be-detected cloth sample into a pre-trained lightweight generative adversarial network, and obtaining a reconstruction image of the to-be-detected cloth sample; s2, comparing the difference between the to-be-detected cloth sample and the reconstructed image of the to-be-detected cloth sample so as to detect cloth defects; wherein the lightweight generator comprises an encoder and a decoder; respectively splicing the feature maps with the same length and width obtained in the encoder and the decoder in a jump connection mode; the encoder comprises a convolutional layer and a bottleneck module which are connected in series; the bottleneck module comprises a plurality of bottleneck layers which are sequentially connected in series from large to small in size; the bottleneck layer comprises a standard convolutional layer, a deep convolutional layer, a channel attention module and a point convolutional layer which are connected in series; the decoder also comprises the bottleneck layer, and the bottleneck layer in the decoder and the bottleneck layer in the encoder are in mirror symmetry; according to the invention, cloth defect detection can be accurately carried out with relatively high operation efficiency.

Description

technical field [0001] The invention belongs to the field of computer vision, and more particularly, relates to a cloth defect detection method and system. Background technique [0002] In the production process of cloth, due to the failure of textile equipment, operator errors and other reasons, the cloth has different forms of defects, which seriously affects the quality of the cloth. At present, most small and medium-sized cloth enterprises still rely on the workers to judge the defects based on their personal experience, so that the quality of the related products cannot be guaranteed. Therefore, there is an urgent need for an effective automated fabric defect detection method. [0003] At present, cloth defect detection methods can be mainly divided into statistical methods, frequency domain methods, model methods and deep learning methods. Among them, the methods based on deep learning have strong robustness to different application scenarios, and have gradually beco...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06T7/00G06N3/04
CPCG06T7/0004G06T2207/30124G06N3/045Y02P90/30
Inventor 孙志刚金栋田莉华肖力王卓
Owner HUAZHONG UNIV OF SCI & TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com