An Infrared and Visible Image Fusion Method Based on Improved GAN with Dropout Layer

In the power grid system, it is of great significance to timely detect the damage and abnormal heating in power equipment as well as maintain the normal and safe operation of the power grid. So far, the usage of infrared and visible image fusion technology to monitor power equipment has low comprehensive expense and high precision so that it has attracted widespread concern. To this end, this paper proposes an improved infrared and visible light GAN image fusion model based on the Dropout layer, which can effectively address the issue of poor generator performance caused by overfitting of the GAN discriminator, without increasing memory usage and training time. Compared to the previous version of the GAN model, it can effectively enhance the quality of the generated fusion image.