Research Article
SAR Moving Target Segmentation and Removal Based on Deep Learning
@INPROCEEDINGS{10.1007/978-3-031-71716-1_9, author={Yifan Wu and Xiyu Qi and Lijia Huang and Bingchen Zhang and Lili Yan}, title={SAR Moving Target Segmentation and Removal Based on Deep Learning}, proceedings={Machine Learning and Intelligent Communication. 8th EAI International Conference, MLICOM 2023, Beijing, China, December 17, 2023, Proceedings}, proceedings_a={MLICOM}, year={2024}, month={9}, keywords={Synthetic Aperture Radar (SAR) Moving target segmentation Deep learning}, doi={10.1007/978-3-031-71716-1_9} }- Yifan Wu
Xiyu Qi
Lijia Huang
Bingchen Zhang
Lili Yan
Year: 2024
SAR Moving Target Segmentation and Removal Based on Deep Learning
MLICOM
Springer
DOI: 10.1007/978-3-031-71716-1_9
Abstract
Synthetic Aperture Radar (SAR) stands as an integral part of advanced remote sensing technology. Nevertheless, practical applications experience inevitable disturbances from moving target noise, compromising both image integrity and target detection performance. This paper introduces a pioneering approach reliant on deep learning principles for the elimination of moving target noise within SAR imaging. Firstly, we use the Back-Projection (BP) algorithm to form the foundational images from echo signals. Employing the Unet network, we subsequently acquire a segmentation map of the moving target noise. By subtracting this segmented noise map from the original image, we succeed in the effective erasure of moving targets, yielding a resultant image devoid of moving target noise. Experimental validation demonstrates that our method can effectively remove moving target noise and yield SAR images that solely contain static scenes.
