The Role of Artificial Intelligence in Enhancing the Accuracy of Visual Inspections: A Neural Network-Based Approach
The Role of Artificial Intelligence in Enhancing the Accuracy of Visual Inspections: A Neural Network-Based Approach
ZAHRA FOROUTANFAR1
1) Department of Chemistry, Shahreza Branch, Islamic Azad University, Shahreza, Iran
IOI Code
:
ABEB-ACZGEF
Publication :
4th International Conference on Recent Advances in Engineering, Innovation and Technology - Belgium(eitconf.com)
Abstract :
Visual inspection of polymer parts is a vital process in quality assurance; however, this process remains highly vulnerable to human error and unstable environmental conditions. This research proposes an advanced automated inspection framework based on deep learning architectures to enhance detection accuracy and operational efficiency. A specialized Convolutional Neural Network (CNN) model was developed and rigorously tested under varying lighting conditions and different noise levels. Experimental results demonstrate that the proposed system achieved an accuracy of 98.2% and an F1-Score of 98.1%, significantly outperforming human inspection and traditional machine vision methods. From a management perspective, the implementation of this AI-based approach led to an 82.2% reduction in scrap rates and a 70.8% reduction in total inspection costs, while increasing production output capacity by more than sixfold. These findings indicate that integrating deep learning into polymer manufacturing processes is a crucial step toward achieving Zero Defect Manufacturing (ZDM) and enhancing global competitiveness within the framework of Industry 4.0.
Keywords : deep learning visual inspection polymer defect detection industry 4 0 zero defect manufacturing quality management