Location and identification of suspension insulators based on RRPN
Chang Wang, Lin Yang, Xiangyu Chen, Shouqiang Fu · 2022 2nd International Conference on Consumer Electronics and Computer Engineering (ICCECE) · 2022
In order to improve the refinement of insulator detection, a suspension insulator location and identification method based on RRPN is proposed in this paper. The RRPN algorithm is used to extract and train the image data of suspension insulator, and the positioning recognition based on inclined box is realized. This paper compares the training curves of the model under different learning rates and selects the optimal learning rate. The average accuracy of the optimal model was 88.47%. The results show that the method proposed in this paper can be combined with the application scenario of suspension insulator UV diagnosis to reduce false detection. The model shows strong robustness in the case of occlusion and truncation and has certain academic innovation and engineering application value.