Deep learning recognition method for ferrite defects based on Fisher criterion

A technology of learning recognition and defect depth, which is applied in the field of surface defect recognition of rectangular ferrite magnetic sheets, can solve the problems of low efficiency and high error rate, and achieve the effect of strengthening detection speed, high accuracy and improving detection accuracy

Inactive Publication Date: 2018-06-05
ZHEJIANG UNIV OF TECH
View PDF3 Cites 6 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

This manual method is inefficient and has a high error rate due to eye strain

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
  • Deep learning recognition method for ferrite defects based on Fisher criterion
  • Deep learning recognition method for ferrite defects based on Fisher criterion
  • Deep learning recognition method for ferrite defects based on Fisher criterion

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0057] The present invention will be further described below in conjunction with the accompanying drawings.

[0058] A deep learning recognition method for ferrite defects based on Fisher's criterion includes the following steps:

[0059] Step 1: Obtain images of rectangular ferrite magnetic sheets of known defects and defect-free samples through an industrial camera;

[0060] Step 2: Obtain images of defective and non-defective ferrite samples;

[0061] Step 3: Define the noise reduction and denoising autoencoder SDA;

[0062] The specific processing method is as follows:

[0063] (1) Define the vectors z and y respectively, and the calculation formula is as follows:

[0064] z=s(Wx+b)

[0065] y=g θ '(z)=s(W'z+b)

[0066] Among them, W is the weight matrix, b is the input bias, s() is the sigmoid function, z is a d'-dimensional vector, x is a d-dimensional input vector, y is also a d-dimensional vector as an approximate vector of x, and the x and y vector elements are ...

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 deep learning recognition method for ferrite defects based on a Fisher criterion. The deep learning recognition method comprises the steps of firstly acquiring images of defect known and defect-free sample rectangular ferrite magnetic pieces through an industrial camera; acquiring images of defective and defect-free ferrite samples; defining a noise reduction automatic encoder (SDA); dividing a ferrite image into patches of the same size, and training an FCSDA by using the defective and defect-free samples; acquiring a positive sample image and a negative sample image of the rectangular ferrite magnetic pieces through the industrial camera; training each DA under an unsupervised mode, wherein the weight and deviation parameters of the DA are used for initializingthe FCSDA rather than a random value; perform fine adjustment on the FCSDA through supervised learning of a labeled data set; training the FCSDA; and dividing test patches into defective test patchesand defect-free test patches through the FCSDA.

Description

technical field [0001] The invention relates to a surface defect identification method of a rectangular ferrite magnetic sheet. Background technique [0002] Defect detection is very important for quality control of ferrite magnets. Traditionally, the human eye detects defects. This manual method is inefficient and has a high error rate due to eye strain. Therefore, it is an urgent need for current production enterprises to develop a high-efficiency and high-precision method for identifying dents on the surface of ferrite wafers. From the current point of view, deep learning is an extremely popular technology, and the defects of magnetic disks can be analyzed and identified through deep learning processing. It makes it possible to improve the accuracy and speed of defect recognition and avoid manual errors. Contents of the invention [0003] The present invention will overcome the above-mentioned defective of prior art, proposes a kind of stack denoising autoencoder FC...

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): G06T7/00G06K9/62
CPCG06T7/0004G06T2207/10004G06T2207/20081G06T2207/30164G06F18/21322G06F18/21324G06F18/24G06F18/214
Inventor 姚明海叶耀威
Owner ZHEJIANG UNIV OF TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap