Bolt Defect Detection Based on CenterNet Model
Nan Yao, Yuxi Zhao, Xi Wu, Ziquan Liu, Jianhua Qin · 2021 IEEE Intl Conf on Dependable, Autonomic and Secure Computing, Intl Conf on Pervasive Intelligence and Computing, Intl Conf on Cloud and Big Data Computing, Intl Conf on Cyber Science and Technology Congress (DASC/PiCom/CBDCom/CyberSciTech) · 2021
Bolts are used to connect various components in power electricity substation. Once lost, it will cause disasters. Therefore, accurate detection of bolt loss is a very important task. Generally speaking, the bolts are small and numerous. If manually inspected, the workload is huge and easy to miss. In this paper, an anchor-free target detection method is proposed to detect the bolt defect. This paper uses CenterNet as a standard to conduct a comparative experiment, and the results show that the method in this paper is significantly improved. In the experiment, the accuracy and callback rate of CenterNet are increased by 8.7% and 4.2% respectively compared with other methods.