Nuts&bolts: YOLO-v5 and image processing based component identification system
Published in Engineering Applications of Artificial Intelligence, 2022
This paper presents a deep learning and image processing approach for identifying mechanical fasteners in aerospace assembly lines. YOLO-v5 for component classification based on head and lateral shape is used and achieved 99.6% mean average precision. Also developed an image processing method to measure fastener dimensions, including thread pitch, with high accuracy (100% for standard sizes, max 0.05 mm error for pitch).
Recommended citation: Mushtaq, Faisel, Kaki Ramesh, Sandip Deshmukh, Tathagata Ray, Chandu Parimi, Praveen Tandon, and Pramod Kumar Jha. "Nuts&bolts: YOLO-v5 and image processing based component identification system." Engineering Applications of Artificial Intelligence 118 (2023): 105665.
Download Paper