A Ship Detector Design Based on Deep Convolutional Neural Networks for Satellite Images

Abstract

Ship target classification from satellite images is a challenging task with its requirements of feature extracting, advanced pre-processing, a variety of parameters obtained from satellites and other type of images, and analyzing of images. The dissimilarity of results, enhanced dataset requirement, intricacy of the problem domain, general use of Synthetic Aperture Radar (SAR) images and problems on generalizability are some topics of the issues related to ship target detection. In this study, we propose a deep convolutional neural network model for detecting the ships using the satellite images as inputs. Our model has acquired an adequate accuracy value by just using a pre-processed satellite image input. Visual and graphical results of features at various layers and deconvolutions are also demonstrated for a better understanding of the basic process.