Research On Identification And Location Of Aerial Insulator Based On Convolutional Neural Network

At present,the regular inspections of high-voltage transmission lines are mainly manual inspections.Staff are required to conduct visual inspections of the transmission lines and even need to climb the iron tower for inspection,which is inefficient and unsafe.With the rapid development of digital image technology and drone technology,power grid companies have proposed intelligent power inspection based on drones,of which insulator identification and positioning based on aerial images is an important part of it.Due to the complex background of the image data collected by the drone and the influence of various factors such as light,the traditional insulator identification and positioning method has low accuracy and slow speed.In view of these problems,this paper conducts research on aerial insulator recognition and location based on convolutional neural network,builds an aerial insulator recognition and location model and compares it with traditional algorithms.Experimental results show that the accuracy rate obtained by the model based on convolutional neural network Significantly higher than traditional algorithms.The main research work of this article is:1.An aerial insulator image data set is constructed,which provides the basis for insulator insulator image research.2.Experiment of insulator image recognition and classification based on traditional algorithm.In the image pre-processing stage,median filtering and edge enhancement are used,then HOG features and Haar features are extracted,and finally SVM classifier and AdaBoost classifier are used to implement insulator image recognition classification.3.The structure and principle of the convolutional neural network are studied,and the target detection network based on the convolutional neural network is studied in detail.The residual module in the ResNet network,the RPN network in the Fast RCNN,and the FPN module are analyzed.Based on the study of the structure and principle of the convolutional neural network,an aerial insulator recognition and localization model based on the convolutional neural network was built,and the border regression algorithm and training optimization algorithm were optimized.The final model achieved an accuracy of 91.3%.Significantly higher than traditional algorithms.