Insulator image data enhancement based on BIG-GAN

Abstract

The data set of power grid equipment is difficult to obtain on a large scale due to the complex working environment of the power grid and many high-altitude operation scenarios, resulting in scarce data in many scenarios. Aiming at the problem of lack of image data of insulators on transmission lines, a data augmentation network based on BIG-GAN network is proposed for the first time. This network combines convolutional neural networks with batch standardization and sampling truncation techniques added to the convolutional neural network, which effectively improve the stability of the GAN network training process and the reconstruction effect of the generative model.