Abstract
Land cover classification with SAR images mainly focuses on the utilization of fully polarimetric SAR (PolSAR) images. The conventional task of PolSAR classification is single-pixel-based region-level classification using polarimetric target decomposition. In recent years, a large number of high-resolution SAR images have become available, most of which are single-polarization. This article explores the potential of object-level semantic segmentation of high-resolution single-pol SAR images, in particular tailored for the Gaofen-3 (GF-3) sensor. First, a well-annotated GF-3 segmentation dataset “FUSAR-Map” is presented for SAR semantic segmentation. It is based on four data sources: GF-3 single-pol SAR images, Google Earth optical remote sensing images, Google Earth digital maps, and building footprint vector data. It consists of 610 high-resolution GF-3 single-pol SAR images with the size of 1024 × 1024. Second, an encoder-decoder network based on transfer learning is employed to implement semantic segmentation of GF-3 SAR images. For the FUSAR-Map dataset, an optical image pretrained deep convolution neural network (DCNN) is fine-tuned with the SAR training dataset. Experiments on the FUSAR-Map dataset demonstrate the feasibility of object-level semantic segmentation with high-resolution GF-3 single-pol SAR images. Also, our algorithm obtains fourth place about the PolSAR image semantic segmentation on the “2020 Gaofen Challenge on Automated High-Resolution Earth Observation Image Interpretation.” The new dataset and the encoder-decoder network are intended as the benchmark data and baseline algorithm for further development of semantic segmentation with high-resolution SAR images. The FUSAR-Map and our algorithm are available at github.com/fudanxu/FUSAR-Map/.
Highlights
S YNTHETIC Aperture Radar (SAR) can obtain rich information of earth surface under all-time and all-weather conditions
We present a pixel-labeled land cover semantic segmentation dataset of GF-3 single-pol SAR images, which is named as FUSAR-Maps
Baseline segmentation algorithm for Gaofen-3 SAR image: Based on the large-scale land cover semantic segmentation dataset of GF-3 single-pol SAR images and the polarimetric SAR (PolSAR) training datasets from “2020 Gaofen Challenge on Automated HighResolution Earth Observation Image Interpretation,” we proposed a unified SAR data preprocessing method for GF-3 SAR data, and a deep learning model using encoder–decoder structure based on transfer learning to achieve object-level semantic segmentation for GF-3 images
Summary
S YNTHETIC Aperture Radar (SAR) can obtain rich information of earth surface under all-time and all-weather conditions. With the rapid development of deep learning method, region-level land cover classification of PolSAR images have emerged, which takes one image patch as input and utilizes convolutional neural networks (CNNs) to extract high-level features and classify the terrain surface [8], [9]. These region-level classification methods can improve the land cover classification performance without any hand-crafted features, its accuracy is still on the level of regional mapping applications. To achieve object-level land cover classification with high-resolution SAR images, which is known as image semantic segmentation in computer vision, deep learning-based segmentation methods have been widely studied [3], [10]–[12]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
