Design and Implementation of Marine Datasets Based on DCGAN Network

In this paper, the deep learning method is applied to the ship detection SAR datasets to enlarge the existing SAR images of ships.This paper first introduces the basic knowledge of convolutional neural networks, and also introduces the common convolutional neural network structure.Secondly, this paper introduces the generative adversarial network GAN,, analyzes the GAN principle and network structure according to the objective function, introduces the training method of the GAN model, and analyzes the advantages and disadvantages of GAN.This introduces the DCGAN network model that combines the above two, analyzes its basic knowledge and network structure, and the improvement of GAN model, and introduces the training process in detail.Finally, we compare the loss function images generated by DCGAN and GAN and summarize the lessons.