TY - GEN
T1 - Faster r-cnn based fault detection in industrial images
AU - Saeed, Faisal
AU - Paul, Anand
AU - Rho, Seungmin
N1 - Publisher Copyright:
© Springer Nature Switzerland AG 2020.
PY - 2020
Y1 - 2020
N2 - 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.
AB - 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.
KW - Convolution neural networks
KW - Defect detection
KW - Fast R-CNN
KW - Fault identification
KW - Industrial images
KW - RPN
UR - http://www.scopus.com/inward/record.url?scp=85091266559&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-55789-8_25
DO - 10.1007/978-3-030-55789-8_25
M3 - Conference contribution
AN - SCOPUS:85091266559
SN - 9783030557881
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 280
EP - 287
BT - Trends 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
A2 - Fujita, Hamido
A2 - Sasaki, Jun
A2 - Fournier-Viger, Philippe
A2 - Ali, Moonis
PB - Springer Science and Business Media Deutschland GmbH
T2 - 33rd International Conference on Industrial, Engineering and Other Applications of Applied Intelligent Systems, IEA/AIE 2020
Y2 - 22 September 2020 through 25 September 2020
ER -