Faster r-cnn based fault detection in industrial images

Faisal Saeed, Anand Paul, Seungmin Rho

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

4 Scopus citations

Abstract

Industry 4.0 requires smart environment to find defects or faults in their products. A defective product in the market can impact negatively on the overall image of the industry. Thus, there is continuous struggle for industrial environment to reduce impulsive downtime, concert deprivation and safety risks. Defect detection in industrial products using the images is very hot topic in era of current research. Machine learning provides various solution but most of the time such solutions are not suitable for environment where product is on conveyor belt and traveling from one point to another. To detect fault using industrial images, we proposed a method which is based on Faster R-CNN which is suitable for smart environment as it can the product efficiently. We simulated our environment using python language and proposed model has almost 99% accuracy. To make our proposed scheme adaptable for the industry 4.0, we also developed an android application which make it easy to interact with the model and industry can train this model according to their needs. Android application is able to take pictures of defective product and feed it to model which improve accuracy and eventually reduces time identify defective product.

Original languageEnglish
Title of host publicationTrends in Artificial Intelligence Theory and Applications. Artificial Intelligence Practices - 33rd International Conference on Industrial, Engineering and Other Applications of Applied Intelligent Systems, IEA/AIE 2020, Proceedings
EditorsHamido Fujita, Jun Sasaki, Philippe Fournier-Viger, Moonis Ali
PublisherSpringer Science and Business Media Deutschland GmbH
Pages280-287
Number of pages8
ISBN (Print)9783030557881
DOIs
StatePublished - 2020
Event33rd International Conference on Industrial, Engineering and Other Applications of Applied Intelligent Systems, IEA/AIE 2020 - Kitakyushu, Japan
Duration: 22 Sep 202025 Sep 2020

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume12144 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference33rd International Conference on Industrial, Engineering and Other Applications of Applied Intelligent Systems, IEA/AIE 2020
Country/TerritoryJapan
CityKitakyushu
Period22/09/2025/09/20

Keywords

  • Convolution neural networks
  • Defect detection
  • Fast R-CNN
  • Fault identification
  • Industrial images
  • RPN

Fingerprint

Dive into the research topics of 'Faster r-cnn based fault detection in industrial images'. Together they form a unique fingerprint.

Cite this