Defect detection algorithm based on deep neural network Mask R-CNN

A deep neural network and defect detection technology, applied in the field of defect detection algorithms based on deep neural network MaskR-CNN, can solve the problems of difficult and accurate segmentation, complex surface texture, low contrast, etc., to achieve accurate defect segmentation and strong generalization ability and robustness effects

Pending Publication Date: 2020-11-24
HUNAN INSTITUTE OF SCIENCE AND TECHNOLOGY +1
View PDF0 Cites 7 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The magnetic tile image has the characteristics of uneven illumination, complex surface texture, and low

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
  • Defect detection algorithm based on deep neural network Mask R-CNN
  • Defect detection algorithm based on deep neural network Mask R-CNN
  • Defect detection algorithm based on deep neural network Mask R-CNN

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0041] see Figure 1-2 , a defect detection algorithm based on deep neural network Mask R-CNN, comprising the following steps:

[0042] S1, utilizing the feature pyramid network (FPN) based on ResNet50 to extract features;

[0043] S2. Using the Region Proposal Network (RPN) to extract the Region of Interest (ROI) of the defect region to obtain the corresponding anchor frame;

[0044] S3. Using a fully convolutional neural network (FCN) to predict the pixel category inside the ROI to achieve defect segmentation;

[0045] S4. Finally, the category of each ROI and the coordinates of the corresponding anchor frame are predicted through the fully connected layer of the network.

[0046] The loss function Loss of Mask R-CNN consists of three parts,

[0047] Loss=L cls +L box +L mask #(1).

[0048] Among them, L cls is the classification loss, L box is the bounding box regression error, L mask is the segmentation loss of the branch FCN.

[0049]FPN is to classify the laye...

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 discloses a defect detection algorithm based on a deep neural network Mask R-CNN, and belongs to the technical field of defect detection. The algorithm comprises the following specific steps of extracting features by using a feature pyramid network (FPN) based on ResNet50; extracting regions of interest (ROI) of a defect region by using a region proposal network (RPN) so as to obtaincorresponding anchor boxes; using a full convolutional neural network (FCN) to predict a pixel category in the ROI so as to realize defect segmentation; and finally, realizing prediction of the category to which each ROI belongs and corresponding anchor frame coordinates through a full connection layer of the network. Aiming at a magnetic shoe surface defect detection scene, the algorithm performs two improvements on a feature pyramid network (FPN) in MaskR-CNN: a C1 module is added in the FPN, and a pooling layer in a feature extraction layer of the C1 module is cancelled; and a CLAHE preprocessing module is added in front of a feature extraction layer of the FPN. Experimental results show that the algorithm of the invention has strong generalization ability and robustness, and can perform accurate defect segmentation on a magnetic shoe image.

Description

technical field [0001] The present invention relates to the technical field of defect detection, more specifically, to a defect detection algorithm based on deep neural network Mask R-CNN. Background technique [0002] With the proposal of "Made in China 2025" strategy and "Industry 4.0", China's traditional manufacturing industry is facing the huge challenge of industrial transformation and industrial upgrading, which promotes the development of products such as industrial robots, high-precision CNC machine tools and new energy vehicles And a wide range of applications, higher performance indicators are also proposed for the motor, and the surface quality of the magnetic tile directly affects the performance of the motor. Magnetic tile surface defect detection technology has the advantages of high detection efficiency, low cost, and high reliability. It is of great significance to the production of motors, and it also promotes the survival and development of enterprises. ...

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): G06T7/00G06T7/11G06N3/08G06N3/04
CPCG06T7/0004G06T7/11G06N3/08G06T2207/20016G06T2207/20081G06T2207/20084G06T2207/20132G06T2207/30108G06N3/045
Inventor 郭龙源段厚裕周武威欧先锋张国云吴健辉鲁敏滕书华
Owner HUNAN INSTITUTE OF SCIENCE AND TECHNOLOGY
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