In view of the increasing demand for power consumption in the whole society,the high cost of power trans-mission,power safety and human maintenance caused by it,this paper uses image enhancement technology and deep learning technology to realize the recognition and analysis of power equipment images.Firstly,image enhancement technolo-gy is used to expand and enhance the power equipment image set.Secondly,a MobileNet convolutional neural network is built to complete data training and recognition of power equipment images.The final experiment shows that MobileNet runs at a speed of 0.02 seconds per image,and the detection rate can reach 96％.This method is suitable for the detection scene of electrical equipment with obvious image features and a large number of repeated detections.
万方数据
维普
