Sentinel-2 Research on the Detection and Classification Methods of Maritime Ship Targets from Remote Sensing Images

Abstract There are problems such as low recognition accuracy and large classification error in the existing methods for ship identification based on optical remote sensing images. In this paper, we will analyze characteristics of ships determine indicative factors applying to monitor combination with Using image data, combined U-Net AttU-Net deep neural network models, assist extracting new indices strong generality clear physical meaning, establishing rules monitoring ships, so establish a more general meaning method satellite The is applied evaluated port data. data show that compared traditional machine learning methods, using models paper reaches 89.04%, recall rate better than SVM. it shows model can detect effectively.