SAR Image Classification of Vessel Based on Data Augmentation and Deep Learning

Ship SAR image classification is one of the most difficult problems in the field of ship target research. The traditional method usually adopts manual processing, which is no longer applicable in the information age when SAR images are more and more abundant. This paper studies the application of convolutional neural network AlexNet model in ship SAR image classification. On the basis of the existing data set, aiming at the problem that the classification accuracy is not high due to the small amount of data, the horizontal flip, vertical flip and shift operations are carried out on the basis of the original image through the data enhancement technology, and the original 3024 images are expanded to 12096 images to meet the requirements of model training. The experimental results show that the classification accuracy is improved to 87.2% after 50 iteration training.