Screen defect classification method, electronic equipment and storage medium

A defect classification and defect type technology, used in instruments, character and pattern recognition, computer parts and other directions, can solve the problems of large image area span, increased maintenance workload, low screen defect classification accuracy, etc., to improve the classification accuracy. rate effect

Active Publication Date: 2022-02-08
高视科技(苏州)股份有限公司
View PDF7 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

When the convolutional neural network algorithm collects the defect ROI area, it usually adopts the sliding window acquisition method, and the image area corresponding to some defects has a large span, which may lead to incomplete defect ROI areas collected, resulting in low defect classification accuracy on the screen
[0003] Traditional image processing algorithms increase the maintenance workload, and the defect ROI area collected by the convolutional neural network algorithm may be incomplete, resulting in low accuracy of screen defect classification

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
  • Screen defect classification method, electronic equipment and storage medium
  • Screen defect classification method, electronic equipment and storage medium
  • Screen defect classification method, electronic equipment and storage medium

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0060] Traditional image processing algorithms need to adjust parameters and operation steps according to different defect types, which increases the workload of maintenance. However, when the convolutional neural network algorithm acquires image defect ROI areas, when the defect ROI area has a large span, it may lead to the collection of defects. The incomplete ROI area will reduce the classification accuracy of the screen defect detection network.

[0061] For the above problems, the embodiment of this application provides a solution, please refer to figure 1 , figure 1 It is a schematic flowchart of defect classification of screen defect images shown in the embodiment of the present application.

[0062] The screen defect image is preprocessed to obtain a defect sample image, and the defect sample image is input into the screen defect detection network for classification, and the defect type of the screen defect image is obtained.

[0063] The defect classification proces...

Embodiment 2

[0083] In order to enhance the generalization ability of the screen defect detection network and improve the classification accuracy of the screen defect detection network.

[0084] The embodiment of this application provides a training method for the screen defect detection network, please refer to figure 2 , figure 2 It is a schematic flowchart of the training method of the screen defect detection network shown in the embodiment of the present application.

[0085] B1. Preprocessing the input screen defect image to obtain a defect sample image.

[0086] In the embodiment of the present application, the content of step B1 is the same as that of step A1 in the first embodiment above, and will not be repeated here.

[0087] B2. Carrying out sample expansion for defect types lacking defect sample images to obtain defect training images.

[0088] Exemplary: performing one or more of rotation, translation, mirroring, and random cropping on a small number of defect sample imag...

Embodiment 3

[0112] After the defect sample images are classified through the screen defect detection network and the defect types of the screen defect images are obtained, it is necessary to improve the classification accuracy of defect types with low classification accuracy and optimize the screen defect detection network.

[0113] The embodiment of this application provides a method for optimizing the screen defect detection network, please refer to Figure 4 , Figure 4 It is a schematic flowchart of an optimized screen defect detection network shown in the embodiment of the present application.

[0114] C1. Add defect training images corresponding to defect types with low classification accuracy.

[0115] For example, if the defect sample images are classified by the screen defect detection network and it is found that the classification accuracy of the screen defects of the white point-like uneven brightness type is low, correspondingly, the defect training images of the white point...

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 screen defect classification method, electronic equipment and a storage medium. The screen defect classification method comprises the steps: preprocessing an input screen defect image to obtain a defect sample image, wherein preprocessing comprises filtering, threshold segmentation, defect area and edge analysis, scale adjustment and downsampling. The defect sample image comprises a defect ROI region, the size of the defect sample image accords with the input image size of the screen defect detection network, and the defect sample image is classified through the screen defect detection network to obtain the defect type of the screen defect image. According to the method and the device, the defect sample image containing the defect ROI can be quickly and accurately acquired by combining the preprocessing and the screen defect detection network, and the defect sample image with the accurate defect ROI can improve the classification accuracy of the screen defect detection network. In addition, the problem that parameters and operation steps need to be adjusted according to different defect types in a traditional algorithm can be solved.

Description

technical field [0001] The present application relates to the technical field of screen defect detection, in particular to a screen defect classification method, electronic equipment and storage media. Background technique [0002] The screen defect classification method can detect the ROI area of ​​the screen defect, and reduce the over-inspection rate and missed-inspection rate of the product. Existing screen defect classification methods are mainly divided into two categories, one is the traditional image processing algorithm, and the other is the convolutional neural network algorithm. Traditional image processing algorithms adjust algorithm parameters and the order of algorithm steps according to different types of defects, which increases the workload of maintenance. When the convolutional neural network algorithm collects the defect ROI area, it usually adopts the sliding window acquisition method, and the image area corresponding to some defects has a large span, wh...

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): G06V10/20G06V10/26G06V10/764G06V10/774G06V10/80G06K9/62
CPCG06F18/24G06F18/25G06F18/214
Inventor 不公告发明人
Owner 高视科技(苏州)股份有限公司
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