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

Surface scratch rapid detection system based on shallow anisotropic pyramid network

An anisotropic, detection system technology, applied in manufacturing computing systems, biological neural network models, character and pattern recognition, etc., can solve the problems of high prediction accuracy, large similarity of samples, and few samples, and achieves time-consuming training. Less, reducing the amount of parameters, the effect of a small amount of parameters

Pending Publication Date: 2022-07-12
重庆守愚科技有限公司
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] In order to solve the following problems in the scratches on the product surface in the prior art: 1) sample collection is difficult, there are few available samples, and there is a large similarity between samples; 2) the accuracy of detection has a serious dependence on detailed information 3) the prediction accuracy requirement is high, the present invention provides a kind of surface scratch detection system based on shallow anisotropic pyramid network and comprises the following steps:

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
  • Surface scratch rapid detection system based on shallow anisotropic pyramid network
  • Surface scratch rapid detection system based on shallow anisotropic pyramid network
  • Surface scratch rapid detection system based on shallow anisotropic pyramid network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0055] In the present invention, use x∈R w×h and y∈R w×h represents the original image and its corresponding manual annotation results, and w and h represent the length and width of the image, respectively. The segmentation model is represented by M:p=f(θ,x), where θ represents the network parameters, p represents the probability map of the prediction result, and M is the abbreviation of the model. Indicates the prediction result:

[0056]

[0057] The present invention includes the following steps:

[0058] S1 data preprocessing

[0059] The collected original image and the corresponding label are resized to 256×256 size respectively, and then normalized by formula 2.

[0060]

[0061] where μ and σ are the mean and variance of the original image data, respectively.

[0062] S2 data augmentation

[0063] The defect collection of product surface scratches is difficult, and there are few data sets available for model training, so it is necessary to enhance the imag...

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 provides a surface scratch rapid detection system based on a shallow anisotropic pyramid network. The surface scratch rapid detection system comprises the following steps: S1, data preprocessing; s2, data enhancement; s3, building a model; s4, model training; the method has the beneficial effects that 1, the method has relatively low network parameter quantity; and 2, feature fusion is carried out by adopting an addition method in Dense connection, and compared with the traditional Dense connection, the parameter quantity is small, and the training time consumption is less, and the anisotropic convolution operation is adopted to replace the traditional convolution operation, so that the parameter quantity of the network is further reduced.

Description

technical field [0001] The invention relates to the field of surface scratch detection, in particular to a surface scratch quick detection system based on a shallow anisotropic pyramid network. Background technique [0002] Scratches on the surface of the product not only affect the appearance of the product, but also have an important impact on its performance (such as lubrication performance, sealing performance, load-bearing performance, etc.), so the accurate detection and identification of surface scratches is an important part of product quality inspection. . At present, the product quality inspection of most enterprises relies heavily on manual labor. This method is time-consuming and labor-intensive, and the test results are easily affected by human factors. In addition, the manual quality inspection method cannot effectively and timely summarize the relevant scratches. It is difficult to play a timely guiding role in the follow-up improvement of the production line...

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): G06K9/62G06N3/04G06N3/08G06V10/80G06V10/764G06V10/774G06V10/82
CPCG06N3/08G06N3/045G06F18/241G06F18/253G06F18/214Y02P90/30
Inventor 周吉祥王翊张勇
Owner 重庆守愚科技有限公司
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