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

Weld defect identification and positioning method and system based on deep learning network

A deep learning network, defect recognition technology, applied in character and pattern recognition, biological neural network model, image data processing and other directions, can solve the problems of low accuracy, insufficient recognition and positioning accuracy, low detection efficiency, etc., to improve the accuracy rate and speed, improve engineering application value, and improve the effect of identification and positioning accuracy

Pending Publication Date: 2021-06-25
TAIYUAN UNIVERSITY OF SCIENCE AND TECHNOLOGY
View PDF6 Cites 19 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] In order to overcome the problems in the prior art, the present invention provides a method for identifying and locating weld defects based on a deep learning network. The disadvantage of low 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
  • Weld defect identification and positioning method and system based on deep learning network
  • Weld defect identification and positioning method and system based on deep learning network
  • Weld defect identification and positioning method and system based on deep learning network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0082] Such as figure 1 As shown, this embodiment provides a method for identifying and locating weld defects based on a deep learning network. The method includes the following steps:

[0083] S1: Obtain radiographic images of weld defects containing pores, slag inclusions, cracks, incomplete fusion and incomplete penetration defects, and use some of the images as training data sets, and the rest as test data sets.

[0084] S2: Perform normalized preprocessing on the images in the test data set or the training data set, after the preprocessing, a block with uniform resolution of the weld defect image is obtained;

[0085] The image preprocessing method is: cut the collected original image according to the specification of 320×320 pixels, and generate the image as an input block. For the original image whose width and height are not an integral multiple of 320×320, the overlap in the graphics is partially reserved. The cutting is done in the way of regions, so that all the cu...

Embodiment 2

[0146] This embodiment provides a simulation experiment of the method for identifying and locating weld defects based on a deep learning network as in Embodiment 1 (in other embodiments, the simulation experiment may not be performed, and other experimental schemes may also be used to conduct experiments to determine the method. The effect of the network model and its related parameters on the weld recognition and localization performance of the method).

[0147] (1) Experimental conditions

[0148] In this embodiment, the detection experiment uses the operating system Windows10, CPUi7-8700k, GPU is 1080ti, memory is 16GB, and the deep learning framework is tensorflow.

[0149] Initialization parameter settings for network training: maximum iteration 50,000 times, learning rate 0.001, batch_size setting 32, weight attenuation coefficient 0.0005, impulse constant 0.9, according to the trend of loss decline, adjust the learning rate and batch_size values ​​appropriately until th...

Embodiment 3

[0183] This embodiment provides a system for identifying and locating weld defects based on a deep learning network. The system uses the method for identifying and locating weld defects based on a deep learning network as in Embodiment 1 to complete the identification of weld defects in radiographic images of welds. and positioning, and give prediction results; the system includes: image acquisition module 1, image preprocessing module 2, and identification and positioning network module 3.

[0184] The image acquisition module 1 is used to acquire radiographic images of weld defects containing pores, slag inclusions, cracks, incomplete fusion and incomplete penetration defects, and use the images as a training set or test set, and complete the alignment based on the images in the training set or test set System training, or complete the task of identifying and locating weld defects in images.

[0185] The image preprocessing module 2 is used to perform normalized preprocessin...

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 the field of image recognition, in particular to a weld defect recognition and positioning method and system based on a deep learning network. The method comprises the following steps: S1, acquiring a weld defect ray image, taking a part of the image as a training data set, and taking the rest image as a test data set; S2, performing normalization preprocessing on the images in the data set; S3, constructing an identifying and positioning network model for identifying and positioning the weld defect image; S4, detecting and processing the weld defect image by using the recognition and positioning network model; S5, training the identification positioning network model until a required training termination condition is reached; And S6, based on the obtained new test data set, identifying and positioning the weld defects by using the trained identifying and positioning network, and evaluating the detection performance of the network module. The method overcomes the defects that a traditional weld defect recognition method is low in accuracy, insufficient in recognition and positioning precision and low in detection efficiency.

Description

technical field [0001] The invention relates to the field of image recognition, in particular to a method and system for identifying and locating weld defects based on a deep learning network. Background technique [0002] Radiographic defect image recognition is an important nondestructive testing method. In the field of ray weld defect image recognition, the manual online detection method is affected by the subjective experience of quality inspectors. When the detection task is large, problems such as missed detection and false detection are likely to occur, which affects the accuracy of the detection results. [0003] In order to solve the above problems, researchers continue to explore the use of artificial intelligence algorithms to automatically identify welding defects in ray weld images. At present, there are two main categories of identification and positioning methods for ray weld defect images based on artificial intelligence algorithms, one is based on tradition...

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
IPC IPC(8): G06T7/00G06K9/32G06K9/62G06N3/04
CPCG06T7/0004G06V10/25G06N3/045G06F18/23213G06F18/253Y02P90/30
Inventor 李砚峰朱彦军孙前来李晔
Owner TAIYUAN UNIVERSITY OF SCIENCE AND TECHNOLOGY
Features
  • Generate Ideas
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More

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

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com