Abstract
Bolts and nuts are the most common parts on automobile sheet metal parts. Due to the complex shape of plate parts, traditional detection methods mostly rely on manual work, and there are problems such as slow detection speed and low accuracy. To solve these problems, this paper uses a target detection method based on YOLOV7 to realize the automatic detection of bolts and nuts on automobile sheet metal parts. By training the automobile sheet metal parts data set for 300 epochs, the final map0.5 reaches 94.89%, and the MAP0.5:0.9 reaches 75.26%. Compared with the test of the model in the coco data set, the map0.5 increases from 69.7% to 94.89%. It shows that the detection effect of the model is very good for this task.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have