CN115222655A - Non-contact rotary knitting yarn detection method - Google Patents

Non-contact rotary knitting yarn detection method Download PDF

Info

Publication number
CN115222655A
CN115222655A CN202210535311.2A CN202210535311A CN115222655A CN 115222655 A CN115222655 A CN 115222655A CN 202210535311 A CN202210535311 A CN 202210535311A CN 115222655 A CN115222655 A CN 115222655A
Authority
CN
China
Prior art keywords
yarn
image
rotary
fiber
knitting
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202210535311.2A
Other languages
Chinese (zh)
Inventor
孙正
单忠德
汪旺
王尧尧
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Nanjing University of Aeronautics and Astronautics
Original Assignee
Nanjing University of Aeronautics and Astronautics
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Nanjing University of Aeronautics and Astronautics filed Critical Nanjing University of Aeronautics and Astronautics
Priority to CN202210535311.2A priority Critical patent/CN115222655A/en
Publication of CN115222655A publication Critical patent/CN115222655A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0004—Industrial image inspection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/11—Region-based segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/136—Segmentation; Edge detection involving thresholding
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20081—Training; Learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20084—Artificial neural networks [ANN]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30108—Industrial image inspection
    • G06T2207/30124—Fabrics; Textile; Paper

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Computational Linguistics (AREA)
  • Evolutionary Computation (AREA)
  • Artificial Intelligence (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Health & Medical Sciences (AREA)
  • Data Mining & Analysis (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Quality & Reliability (AREA)
  • Treatment Of Fiber Materials (AREA)

Abstract

The invention discloses a non-contact rotary weaving yarn detection method, which comprises the following steps: (1) Establishing a storage knowledge base, completing yarn type identification and extracting an original image of a yarn carrier on a rotary knitting machine; (2) Preprocessing the extracted original image and collecting the preprocessed image; (3) Defining the defects of the yarn, including yarn fiber breakage, yarn fiber knotting and yarn fiber crossing; (4) Carrying out characteristic identification on the defects of the yarns according to the preprocessed images; (5) Carrying out yarn detection by using a binocular camera, and analyzing the knitting state of the yarn according to the yarn defect definition; (6) extracting data of the knitting state of the yarn; (7) And if the yarn fiber is broken, the yarn fiber is knotted and the number of crossed yarns is larger than a preset value, stopping the rotary knitting machine. The invention detects the rotary knitting yarn by the image detection technology, effectively improves the detection efficiency, reduces the machine error and improves the product quality.

Description

Non-contact type rotary knitting yarn detection method
Technical Field
The invention relates to the technical field of rotary three-dimensional weaving, in particular to a non-contact rotary weaving yarn detection method in the aspect of rotary three-dimensional weaving yarn detection.
Background
At present, most textile technologies use contact type yarn detectors, a sensor testing element is in direct contact with yarns in the detection process, the surfaces of the detectors are in contact with the moving yarns for a long time, abrasion is generated, the service life of the detectors is shortened, the running state of the yarns is influenced, and the quality and the change of the yarns cannot be truly reflected. The non-contact yarn detection method can obviously reduce the influence of an instrument on the yarn conveying state, thereby improving the accuracy of yarn detection.
Disclosure of Invention
The purpose of the invention is as follows: aiming at the defects of the prior art, the invention provides a non-contact type rotary knitting yarn detection method, which solves the problems of abrasion and inaccuracy of detection of rotary knitting contact type yarns and greatly improves the accuracy of rotary knitting yarn detection.
The technical scheme is as follows: in order to solve the problems, the technical scheme adopted by the non-contact rotary weaving yarn detection method provided by the invention comprises the following steps:
a non-contact rotary weaving yarn detection method is characterized by comprising the following steps:
s1, establishing a reserve knowledge base, completing yarn type identification and extracting an original image of a yarn carrier on a rotary knitting machine: carrying out data processing on the yarn type image and guiding the yarn type image into a reserve knowledge base, and automatically identifying the yarn type by a system; capturing an original image of the movement of the yarn carrier in real time by a camera;
s2, preprocessing the extracted original image and collecting the preprocessed image, wherein the preprocessing comprises the steps of carrying out image numerical processing on the original image through an A/D (analog/digital) converter to obtain a digital image, preprocessing the digital image to remove noise and strengthen image characteristics;
s3, defining yarn defects, including yarn fiber breakage, yarn fiber knotting and yarn fiber crossing;
s4, carrying out characteristic identification on yarn defects in the preprocessed digital image: dividing the preprocessed digital image by adopting a threshold value division method, extracting the characteristics of the divided area, and comparing and identifying the characteristics with the yarn defect characteristics input into a yarn detection system to determine the defect type of the yarn on the yarn carrier;
s5, detecting the yarns by using a camera, and analyzing the knitting state of the yarns according to the yarn defect definition: a camera is arranged at the position of a yarn guide opening of the yarn carrier, and the defect type of the yarn on the yarn carrier is determined by the image shot by the camera through the defect characteristic identification;
s6, extracting data of the knitting state of the yarn: determining the knitting state of the yarn according to the steps, and obtaining the breaking degree of the yarn fibers, the knotting and the crossed number of the yarn fibers through data post-processing;
and S7, if the yarn fiber has section fracture, yarn fiber knotting and the number of crossed fibers is larger than a preset value, stopping the rotary knitting machine.
Further, the step S3 of defining the defect of the yarn further includes: and the neural network based on deep learning memorizes and feeds back the yarn defects.
Further, in the threshold analysis in step S4, that is, if more than half of the volume of the image pixel is air, the yarn defect is identified as yarn breakage, and if the volume of the image pixel is continuously full, the yarn is identified as knotting and crossing.
Further, the extracting data of the knitting state of the yarn in the step S6 further includes: and (3) analyzing and calculating the yarn fiber breaking degree, yarn fiber knotting and crossing number based on the neural network of deep learning.
Further, the stopping of the rotary knitting machine in step S7 further includes: and (3) real-time feedback of the yarn state of the rotary braiding machine, adjusting the chassis motion parameters of the rotary braiding machine and the yarn path plan through yarn state data, realizing self-detection feedback adjustment of the rotary braiding machine, and stopping the rotary braiding machine if the self-detection feedback adjustment exceeds the self-adjustment range.
The invention provides a non-contact rotary knitting yarn detection method. Compared with the prior art, the method has the following beneficial effects:
(1) The non-contact rotary weaving yarn detection method constructs a yarn storage knowledge base and yarn defect type definitions, and solves the non-contact yarn detection problems of complex operation and low detection efficiency.
(2) According to the non-contact rotary weaving yarn detection method, data analysis processing is carried out on the image through a computer vision system and a neural network based on deep learning, and yarn detection of the non-contact rotary weaving machine can be achieved.
Drawings
Fig. 1 is a flow chart of a non-contact rotary knitting yarn detecting method according to the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
As shown in the figure, an embodiment of the present invention provides a technical solution, including the following steps:
s1, establishing a reserve knowledge base, completing yarn type identification and extracting an original image of a yarn carrier on a rotary knitting machine: the yarn type image is subjected to data processing and is led into a storage knowledge base, the system automatically identifies the yarn type, the next action is convenient to carry out, if the yarn cannot be identified, the storage knowledge base is added or the yarn cannot be identified again, and then the binocular camera captures the original image of the motion of the yarn carrier in real time;
s2, preprocessing the extracted original image and collecting the preprocessed image: carrying out image digitization processing on an original image through an A/D converter, and then preprocessing the digitized image to remove noise and strengthen image characteristics;
s3, defining the defects of the yarn, including yarn fiber breakage, yarn fiber knotting and yarn fiber crossing: summarizing and analyzing problems of normal working yarns of a rotary braiding machine, defining defects of the yarns according to literature research, and inputting the defects into a yarn detection system;
s4, carrying out characteristic identification on the defects of the yarns according to the preprocessed image: after preprocessing by adopting a threshold segmentation method, extracting the characteristics of the segmentation area and comparing and identifying the characteristics with the yarn defect characteristics input into a yarn detection system;
s5, carrying out yarn detection by using a binocular camera, and analyzing the knitting state of the yarn according to the yarn defect definition: four binocular cameras are arranged at the yarn guide port of the yarn carrier, and the defect type of the yarn on the yarn carrier is determined by the defect characteristic identification of the image shot by each camera;
s6, extracting data of the knitting state of the yarn: determining the knitting state of the yarn according to the steps, and obtaining the breaking degree of the yarn fibers, the knotting and the crossed number of the yarn fibers through data post-processing;
and S7, if the yarn fibers have more than 50% of broken sections, are knotted and have the number of crossed fibers larger than a preset value, stopping the rotary knitting machine.
In this embodiment, the defining the defect of the yarn in S3 further includes: and the neural network based on deep learning memorizes and feeds back the yarn defects. The neural network based on deep learning is used for constructing a corresponding yarn detection network aiming at the characteristics of yarn fiber image data to obtain a yarn defect detection model; the yarn detection network adopts a characteristic pyramid network, the characteristic pyramid network enhances a convolutional neural network through a top-down path and transverse connection, a multi-scale characteristic pyramid is constructed in an input image acquired by a camera, each layer of the pyramid is used for detecting yarn defects of different scales, and then abnormal results of analysis are fed back in real time. The technical means of applying the characteristic pyramid network to the yarn detection network belongs to the prior art, and reference may be made to chinese patent application CN 113592852A, which is not described herein again.
The threshold analysis in the step S4 further includes a novel 50% threshold analysis method, that is, if more than half of the volume of the image pixel is air, the yarn defect is identified as yarn breakage, and if the volume of the image pixel is continuously full, the yarn is identified as knotting and crossing.
The extracting of the data of the knitting state of the yarn in S6 further includes: and (3) analyzing and calculating the yarn fiber breaking degree, the yarn fiber knotting and the number of crossed yarns based on the neural network of deep learning.
The step S7 of stopping the operation of the rotary braiding machine further includes: and (3) real-time feedback of the yarn state of the rotary braiding machine, adjusting the chassis motion parameters of the rotary braiding machine and the yarn path plan through yarn state data, realizing self-detection feedback adjustment of the rotary braiding machine, and stopping the rotary braiding machine if the self-detection feedback adjustment exceeds the self-adjustment range.

Claims (6)

1. A non-contact rotary weaving yarn detection method is characterized by comprising the following steps:
s1, establishing a reserve knowledge base, completing yarn type identification and extracting an original image of a yarn carrier on a rotary knitting machine: carrying out data processing on the yarn type image and guiding the yarn type image into a reserve knowledge base, and automatically identifying the yarn type by a system; capturing an original image of the movement of the yarn carrier in real time by a camera;
s2, preprocessing the extracted original image and collecting a preprocessed image, wherein the preprocessing comprises the steps of carrying out image numerical processing on the original image through an A/D converter to obtain a digital image, preprocessing the digital image to remove noise and strengthen image characteristics;
s3, defining yarn defects, including yarn fiber breakage, yarn fiber knotting and yarn fiber crossing;
s4, carrying out characteristic identification on yarn defects existing in the preprocessed digital image: segmenting the preprocessed digital image by adopting a threshold segmentation method, extracting the characteristics of the segmented area, and comparing and identifying the characteristics with the yarn defect characteristics input into a yarn detection system to determine the defect type of the yarn on the yarn carrier;
s5, detecting the yarns by using a camera, and analyzing the knitting states of the yarns according to the yarn defect definition: a camera is arranged at the position of a yarn guide opening of the yarn carrier, and the defect type of the yarn on the yarn carrier is determined by the image shot by the camera through the defect characteristic identification;
s6, extracting data of the knitting state of the yarn: determining the knitting state of the yarn according to the steps, and obtaining the breaking degree of the yarn fiber, the knotting of the yarn fiber and the number of crossed yarns through data post-processing; the knitting state comprises whether the yarn knitting is normal or not and whether the yarn defect defined before exists or not;
s7, if the yarn fiber has section fracture, yarn fiber knotting and the number of crossed fibers is larger than a preset value, the rotary braiding machine stops working.
2. The method of claim 1, wherein in step S1, images of different types of original yarns are input into a storage knowledge base, the system captures the images according to their characteristics, and then classifies the yarns according to their characteristics.
3. The method of claim 1, wherein the step S3 of defining the yarn defect further comprises: and the yarn defects are stored in a deep learning neural network, and the yarn defects are memorized and fed back based on the deep learning neural network.
4. The method of claim 2, wherein in step S4, the threshold analysis is performed such that if at least half of the volume of the image pixels is air, the yarn defect is identified as yarn breakage, and if the volume of the image pixels is continuously filled, the yarn is identified as yarn knot and crossing.
5. The method of claim 2, wherein the extracting data of the yarn weaving state in step S6 further comprises: and (3) analyzing and calculating the yarn fiber breaking degree, the yarn fiber knotting and the number of crossed yarns based on the neural network of deep learning.
6. The non-contact rotary knit yarn detection method of claim 4 wherein the step S7 of deactivating the rotary knit machine further comprises: and (3) real-time feedback of the yarn state of the rotary braiding machine, adjusting the chassis motion parameters of the rotary braiding machine and the yarn path plan through yarn state data, realizing self-detection feedback adjustment of the rotary braiding machine, and stopping the rotary braiding machine if the self-detection feedback adjustment exceeds the self-adjustment range.
CN202210535311.2A 2022-05-17 2022-05-17 Non-contact rotary knitting yarn detection method Pending CN115222655A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210535311.2A CN115222655A (en) 2022-05-17 2022-05-17 Non-contact rotary knitting yarn detection method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210535311.2A CN115222655A (en) 2022-05-17 2022-05-17 Non-contact rotary knitting yarn detection method

Publications (1)

Publication Number Publication Date
CN115222655A true CN115222655A (en) 2022-10-21

Family

ID=83608247

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210535311.2A Pending CN115222655A (en) 2022-05-17 2022-05-17 Non-contact rotary knitting yarn detection method

Country Status (1)

Country Link
CN (1) CN115222655A (en)

Similar Documents

Publication Publication Date Title
CN107886500A (en) A kind of production monitoring method and system based on machine vision and machine learning
CN114897894B (en) Method for detecting defects of cheese chrysanthemum core
CN108364291A (en) Grey cloth rapid detection method based on computer vision technique
CN102221559A (en) Online automatic detection method of fabric defects based on machine vision and device thereof
CN113920086B (en) Yarn state detection method and device in carbon fiber weaving process
CN112767384B (en) Compensation method and device of elastic fabric gram weight on-line detection system
CN110567969B (en) Image recognition method and system for detecting fabric defects
CN115063424B (en) Textile bobbin yarn detection method based on computer vision
CN114998321B (en) Textile material surface hairiness degree identification method based on optical means
CN112330598A (en) Method and device for detecting stiff silk defects on chemical fiber surface and storage medium
CN103116351A (en) Spinning defective cloth detection camera and detection system thereof
CN115266732B (en) Carbon fiber tow defect detection method based on machine vision
KR102445162B1 (en) Fabric inspection system
CN110458809B (en) Yarn evenness detection method based on sub-pixel edge detection
CN109030499A (en) A kind of device and method preventing defect counts repeat count suitable for target defect continuous on-line detection
CN110415221A (en) A kind of container truck based on Image Feature Point Matching is anti-to sling automatic testing method
CN115222655A (en) Non-contact rotary knitting yarn detection method
CN116563276B (en) Chemical fiber filament online defect detection method and detection system
CN109632817B (en) Fabric weaving on-line defect detection method based on collimated laser beam
CN110672209A (en) Online detection method for warp knitting cloth color difference
CN203069527U (en) Textile defective cloth detection camera and detection system thereof
CN114782426A (en) Knitted fabric broken yarn defect detection method based on artificial intelligence system
CN113393444B (en) Polyester DTY network point detection method based on image processing technology
CN117273554B (en) Textile production quality prediction method based on data identification
CN117745730B (en) Polyester filament yarn detection method and system based on image processing technology

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination