Power transmission line target detection and identification method based on feature fusion

The invention discloses a power transmission line target detection and identification method based on feature fusion, and the method comprises the steps: S1, constructing a power transmission line image database, and amplifying a database sample; marking the data set image as a VOC format to obtain the frame position and category information of each component in the image; carrying out clustering analysis on the marked image data to obtain prior frame width and height dimension parameters which are used as initial parameters of network training; the YOLOv5 network structure is optimized and adjusted; the pre-training YOLOv5 network is pre-trained, a training set and a verification set of the power transmission line image data set are input into the pre-trained network for continuous training and verification, and an optimal YOLOv5 network model is obtained; and performing target detection and identification on the test set image by using the optimal YOLOv5 network model. According to the method, the position