Abstract Machine Learning and Computer Vision are increasingly being applied in detecting product defects across various industries such as industrial and agricultural, leading to increased efficiency, accuracy, and reduced labor costs. In this study, we utilized image processing algorithms with OpenCV library, combined with deep learning model FASTER R-CNN to identify bearing faults. Unlike previous studies that mainly focused on measuring box-shaped objects or only identifying the outer radius of an object, our study emphasizes identifying the radii of bearings along with a deviation of 0.02 mm. The porposed FASTER R-CNN model to accurately identify faulty bearings with a precision of 98%. Through our research and experimentation, we have also found that the CNN model is more accurate in detection than other models such as YOLO and SSD.
Alan : Mühendislik
Dergi Türü : Uluslararası
Benzer Makaleler | Yazar | # |
---|
Makale | Yazar | # |
---|