Defect Detection Method of Overhead Transmission Line Based on UAV
Aimed at the problems of high data dependence,large amount of calculation and parameters,and limited detection ac-curacy in the current transmission line detection model based on UAV platform,an intelligent transmission line detection meth-od based on UAV is proposed.An improved Faster R-CNN defect detection algorithm is proposed,using MobileNet,soft non-maximum suppression and context-aware ROI pooling layer respectively to improve the model detection accuracy and sensitivity to small size components.In view of the complex noise environment in the image,the Kalman filter algorithm is used to opti-mize and calibrate the detection results,so as to further improve the detection accuracy and stability of the model.The ground high-performance server training and the actual test of the UAV platform are taken as examples to verify the proposed model.The test results show that the proposed model has the best performance compared with YOLO and SSD Fast R-CNN.Through the UAV platform test,the average accuracy of the proposed model is 84.37％,the network scale is 19.8 MB,and the FPS can reach 28.7 Frame/s.
万方数据
维普
