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

Assembly change detection method and device based on attention mechanism and medium

A technology of change detection and attention, applied in neural learning methods, manufacturing computing systems, computer parts, etc., can solve problems such as single color and texture information of parts, large viewing angle changes, and complex structures

Active Publication Date: 2021-08-17
QINGDAO TECHNOLOGICAL UNIVERSITY
View PDF11 Cites 3 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

At present, there are few studies on multi-view change detection of mechanical assemblies. This is mainly because compared with satellite images, mechanical parts have more complex structures, serious occlusions, large changes in viewing angles, and single color and texture information of parts. It is difficult to detect changes in its assembly process, and there is a lack of corresponding data sets

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
  • Assembly change detection method and device based on attention mechanism and medium
  • Assembly change detection method and device based on attention mechanism and medium
  • Assembly change detection method and device based on attention mechanism and medium

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0040] see figure 1 , an attention mechanism-based assembly change detection method, including the following steps:

[0041] Establish a data set; establish a three-dimensional model of a mechanical assembly through SolidWorks software, and add labels to each part in the three-dimensional model. In this embodiment, the labels added to the parts are color marks, and m assembly nodes are set, and m-1 assembly Steps, each step assembles a part, then loads the 3D model of the mechanical assembly into the depth image and color image imaging software, sets the virtual camera to perform imaging processing on different angles of each assembly node, and obtains each assembly node The depth image and color image of the 3D model under different viewing angles, and use the color marking of the color image to generate the change label image of the newly assembled parts of each assembly node;

[0042] Train the detection model; select the depth image of the 3D model of the previous assembl...

Embodiment 2

[0051] see image 3 , this embodiment proposes a specific implementation of the feature extraction module:

[0052] First, perform 3×3 convolution on the input fusion image, and use the four stages of the RepVGG classification network to extract the features of the input image;

[0053] After each stage of the RepVGG classification network, the attention mechanism is embedded to obtain four sets of feature maps. After performing 1×1 convolution on the four sets of feature maps, the channel size of the feature maps is converted to P1, and P1 is set to is defined as the average channel size of the first two sets of feature maps;

[0054] The size of the feature maps of the feature maps of the last three stages is uniformly up-sampled to the size of the feature maps of the first stage, and the obtained four sets of feature maps are fused in series, and the fused feature maps are sequentially processed for 3 ×3 convolution and 1×1 convolution to obtain initial feature maps. Due...

Embodiment 3

[0067] An assembly change detection device based on an attention mechanism, comprising a memory, a processor, and a computer program stored on the memory and operable on the processor, characterized in that, when the processor executes the program, the An assembly change detection method based on an attention mechanism described in any embodiment of the invention.

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 an assembly change detection method based on an attention mechanism, and the method comprises the following steps: building a three-dimensional model of an assembly, adding a label to each part in the three-dimensional model, setting a plurality of assembly nodes, obtaining depth images of the three-dimensional model at different visual angles under each assembly node, obtaining a change label image of a newly added part of each assembly node; selecting two depth images at different visual angles at two moments before and after as training samples; training samples being sequentially subjected to semantic fusion, feature extraction, attention mechanism processing and metric learning, the detection model being trained, the training samples being continuously selected to train the detection model, model parameters with the optimal similarity in the training process being stored, and training being completed; and obtaining depth images of front and rear assembly nodes in the assembly process of the to-be-detected assembly, inputting the depth images into the trained detection model, and outputting a change image of newly added parts of the assembly in the assembly process.

Description

technical field [0001] The invention relates to an assembly multi-angle change detection method based on an attention mechanism, and belongs to the technical fields of computer vision and intelligent manufacturing. Background technique [0002] Computer vision is of great significance to the upgrading of intelligent manufacturing, especially the emergence of a large number of deep learning networks has promoted the development of modern industry. In the mass customization production assembly process, the continuous change of product types will increase the difficulty of assembling products. In the assembly process of complex assembly parts, if the correctness of the newly assembled parts cannot be detected in time, the quality and assembly efficiency of mechanical products will be affected. Therefore, in the assembly process of mechanical assemblies, detecting new assembly parts in each assembly step from multiple perspectives will help to obtain relevant information about ...

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): G06K9/62G06N3/04G06N3/08G06T17/00
CPCG06T17/00G06N3/08G06V10/757G06N3/045G06F18/22G06F18/25G06F18/2415Y02P90/30G06V2201/06G06V10/82G06V20/647G06F18/2148G06F18/213G06F18/217
Inventor 陈成军李长治李东年洪军
Owner QINGDAO TECHNOLOGICAL UNIVERSITY
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