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

Image classification method and device and storage medium

A classification method and image technology, applied in instruments, character and pattern recognition, computer parts, etc., can solve the problem of low detection accuracy of solder joints

Pending Publication Date: 2020-02-04
TP-LINK
View PDF5 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] In view of this, the embodiment of the present invention provides an image classification method, device and storage medium to solve the problem of low detection accuracy of solder joints

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
  • Image classification method and device and storage medium
  • Image classification method and device and storage medium
  • Image classification method and device and storage medium

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0041] Such as figure 1 What is shown is a schematic flowchart of an image classification method provided in Embodiment 1 of the present invention. This embodiment is applicable to the application scenario of defect detection of solder joints on the image to be detected. The method can be executed by a detection device, which can be a server, a smart terminal, a tablet or a PC, etc.; in the embodiment of the invention, The detection device is described as the main body of execution, and the method specifically includes the following steps:

[0042] S110. Segment the original image to obtain several images to be detected;

[0043] The image classification method can be applied to a variety of application scenarios that require image classification. In this embodiment, the application scenario of using the image classification method for solder joint defect detection is described. In order to detect the solder joints on the workpiece or product, Take a picture of the weld on the wor...

Embodiment 2

[0060] Such as image 3 Shown is a schematic flow chart of the image classification method provided in the second embodiment of the present invention. On the basis of Embodiment 1, this embodiment also provides a training process of an image classification model, so as to solve the problem of difficult samples. The method specifically includes:

[0061] S210. Establish a difficult case pool to perform iterative training on the image classification model according to error-prone samples in the difficult case pool.

[0062] After the image classification model is trained, the image classification model can also be iteratively trained for error-prone samples, and oversampling is used for multiple iteration training to solve the problem of difficult samples. Among them, the error-prone samples that are classified incorrectly after each training of the image classification model are stored in the hard case pool. The error-prone samples used in each iteration of training are obtained f...

Embodiment 3

[0067] Such as Figure 4 Shown is the image classification device provided in the third embodiment of the present invention. On the basis of Embodiment 1 or 2, an embodiment of the present invention also provides an image classification device 4, which includes:

[0068] The image segmentation module 401 is used to segment the original image to obtain several images to be detected;

[0069] The calculation module 402 is configured to input the to-be-detected image into an image classification model for calculation for each of the to-be-detected images to obtain first feature data;

[0070] In an implementation example, the calculation module 402 inputs the image to be detected into the image classification model for calculation, and before the first feature data is obtained, the device further includes:

[0071] The difficult case pool establishment module is used to establish a difficult case pool to iteratively train the image classification model according to error-prone samples in...

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 belongs to the technical field of machine vision, and provides an image classification method and device, and a storage medium, and the method comprises the steps: segmenting an originalimage, and obtaining a plurality of to-be-detected images; for each to-be-detected image, inputting the to-be-detected image into an image classification model for calculation to obtain first featuredata; performing image feature extraction on the to-be-detected image, and extracting preset target features to obtain second feature data; combining the first feature data and the second feature data to generate a feature image; and performing image classification on the feature image through the image classification model, and outputting a classification result. According to the embodiment of the invention, the second feature data obtained by extracting the preset target features from the to-be-detected image and the first feature data obtained by convolution are combined to generate the feature map, the feature map is input into the image classification model again for image classification, the network structure of the image classification model is improved, and the problem of low classification accuracy is solved.

Description

Technical field [0001] The present invention relates to the technical field of machine vision, in particular to an image classification method, device and storage medium. Background technique [0002] After wave soldering on the production line, due to the immature soldering process, continuous soldering and missing soldering are often caused, which will greatly affect the service life and appearance of the product, and it needs to be quality tested. In the prior art, the image processing methods of denoising, transformation, segmentation and feature extraction in traditional visual algorithms are usually used to process the solder joint images, and then the solder joints are classified by SVM (Support Vector Machine); or the traditional visual algorithm is used. After the image processing method in the algorithm processes the image of the solder joints, SVM is used to classify the solder joints when the number of solder joints is small, and the CNN convolutional neural network i...

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): G06K9/62
CPCG06F18/24G06F18/253G06F18/254G06F18/214
Inventor 韦奕龙张强王朝允戴盾
Owner TP-LINK
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