Fabric defect detection method based on Swinin-Transform and NAS-FPN

A detection method and defect technology, applied in the field of computer vision, can solve problems such as inapplicability and unrecognizable extremely small targets, and achieve good feature expression and feature fusion effects

Pending Publication Date: 2022-02-18
WUHAN TEXTILE UNIV
View PDF0 Cites 5 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, it is not applicable to directly use the general-purpose target detection method to detect fabrics. This is mainly because the general-purpose target detection model cannot recognize extremely small targets, and the defects on many fabrics are just extremely small, such as stains, thread ends, holes, etc., so the present invention proposes Swin-Transformer and NAS-FPN to solve this problem

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
  • Fabric defect detection method based on Swinin-Transform and NAS-FPN
  • Fabric defect detection method based on Swinin-Transform and NAS-FPN
  • Fabric defect detection method based on Swinin-Transform and NAS-FPN

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0027] The principles and features of the present invention are described below, and the examples given are only used to explain the present invention, and are not intended to limit the scope of the present invention.

[0028] Such as figure 1 As shown, the embodiment of the present invention provides a fabric defect detection method based on Swin-Transformer and NAS-FPN, the specific steps are as follows:

[0029] Step 1. Use the image acquisition system to collect fabric images in the cloth research workshop of the textile enterprise. 4365 images with a resolution of 4096×1696 are collected, and the ratio of the number of images in the training set to the test set is 4:1.

[0030] Table 1 Quantity of various defects

[0031]

[0032] Step 2. Build the Swin-Transformer module, such as figure 2 As shown, the Swin-Transformer module is mainly composed of region division and 4 stages. The region division module divides the input image into non-overlapping small blocks, and...

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 relates to a fabric defect detection method based on Swinin-Transform and NAS-FPN, which mainly comprises three stages: in the first stage, establishing a data set, collecting fabric images with defects, manually marking the positions and types of the defects, and dividing the data set into two parts, namely a training set and a test set; in the second stage, the model is trained, on a training set, input image features are extracted through a Swin-Transform module, then feature fusion is conducted through an NAS-FPN module, and a rough defect position frame is obtained through an RPN module according to the obtained features; and in the third stage, the position and the type of the final fabric defect are obtained by processing the characteristics. Compared with a general target detection network, the detection result of the target detection model constructed by combining the Swinin-Transform and the NAS-FPN is more accurate, and the target detection model is particularly more targeted to fabric defects.

Description

technical field [0001] The invention relates to the technical field of computer vision, and relates to a fabric defect detection method based on Swin-Transformer and NAS-FPN. Background technique [0002] During the production process, due to equipment, raw materials, workmanship or other external factors, the surface of the cloth may have defects. Defects will affect the quality of fabrics, and ultimately affect the economic benefits of enterprises. There are many types of defects, irregular in shape and size; the background texture of the fabric image is very complex, which makes it difficult to realize the automatic detection of fabric defects. Not only that, the lack of large-scale defect image datasets also brings difficulties to the training of defect detection. For the process of defect detection, most companies currently rely on manual visual inspection. This method not only has low efficiency and poor precision, but also causes irreversible damage to the eyes of ...

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 & AuthorityApplications(China)
IPC IPC(8): G06T7/00G06N3/02G06V10/774G06V10/80G06K9/62
CPCG06T7/0004G06N3/02G06T2207/30124G06F18/253G06F18/214
Inventor李敏雷承霖王斌朱萍何儒汉胡新荣
OwnerWUHAN TEXTILE UNIV