Abstract: Sentinel is an AI-powered visual inspection system designed for automatic detection of anomalies and defects in manufactured products, with a particular focus on steel surfaces. Leveraging computer vision and deep learning, Sentinel offers real-time defect detection during manufacturing, identifying anomalies like roll printing, iron-oxide scales, inclusions, scratches, holes, and cracks. By integrating with production line camera feeds, Sentinel provides continuous quality evaluation and improvement. Our approach, utilizing the YOLO (You Only Look Once) model, streamlines the detection process, reducing computational complexity and achieving faster inference speeds. Through extensive experimentation and evaluation on real-world steel defect datasets, our system aims to enhance the efficiency and accuracy of defect detection, paving the way for improved steel quality control, production efficiency, and safety.
Keywords: Manufacturing quality control, YOLO object detection, Early defect identification, Product quality assurance
Cite:
B V Suresh Reddy, Kandula Haripriya, Budda Sivaparvathi, Mutchintala Renuka, Narne Venkata Nagalakshmi, Puvvula Sarayu,"Sentinel - Intelligent Defect Detection System", IJARCCE International Journal of Advanced Research in Computer and Communication Engineering, vol. 13, no. 3, 2024, Crossref https://doi.org/10.17148/IJARCCE.2024.13308.
|
DOI:
10.17148/IJARCCE.2024.13308
