Abstract
SAR Ship Detection Dataset (SSDD) is the first open dataset that is widely used to research state-of-the-art technology of ship detection from Synthetic Aperture Radar (SAR) imagery based on deep learning (DL). According to our investigation, up to 46.59% of the total 161 public reports confidently select SSDD to study DL-based SAR ship detection. Undoubtedly, this situation reveals the popularity and great influence of SSDD in the SAR remote sensing community. Nevertheless, the coarse annotations and ambiguous standards of use of its initial version both hinder fair methodological comparisons and effective academic exchanges. Additionally, its single-function horizontal-vertical rectangle bounding box (BBox) labels can no longer satisfy the current research needs of the rotatable bounding box (RBox) task and the pixel-level polygon segmentation task. Therefore, to address the above two dilemmas, in this review, advocated by the publisher of SSDD, we will make an official release of SSDD based on its initial version. SSDD’s official release version will cover three types: (1) a bounding box SSDD (BBox-SSDD), (2) a rotatable bounding box SSDD (RBox-SSDD), and (3) a polygon segmentation SSDD (PSeg-SSDD). We relabel ships in SSDD more carefully and finely, and then explicitly formulate some strict using standards, e.g., (1) the training-test division determination, (2) the inshore-offshore protocol, (3) the ship-size reasonable definition, (4) the determination of the densely distributed small ship samples, and (5) the determination of the densely parallel berthing at ports ship samples. These using standards are all formulated objectively based on the using differences of existing 75 (161 × 46.59%) public reports. They will be beneficial for fair method comparison and effective academic exchanges in the future. Most notably, we conduct a comprehensive data analysis on BBox-SSDD, RBox-SSDD, and PSeg-SSDD. Our analysis results can provide some valuable suggestions for possible future scholars to further elaborately design DL-based SAR ship detectors with higher accuracy and stronger robustness when using SSDD.
Highlights
Detecting ships in the vast ocean, trade-bustling rivers, and important ports is conducive to traffic control, trade activity monitoring, fishery surveillance, and defense deployment
We expect that this review will be useful for relevant scholars who are studying Synthetic Aperture Radar (SAR) ship detection based on deep learning (DL)
This article reviews the current usage status of the first open SAR Ship Detection Dataset (SSDD) dataset in the SAR ship detection community
Summary
Detecting ships in the vast ocean, trade-bustling rivers, and important ports is conducive to traffic control, trade activity monitoring, fishery surveillance, and defense deployment. Since the United States launched the first SAR satellite on 28 June 1978 [5], a variety of SAR ship detection methods have emerged [6] based, e.g., on constant false alarm rate (CFAR) [7], generalized likelihood ratio test (GLRT) [8], visual saliency [9], super-pixel segmentation [10], polarization decomposition [11], and some auxiliary features (e.g., oil spill clues and ships’ wake) [12,13]. DL provides an increasing number of elegant solutions for various communities, e.g., computer vision (image classification and object detection), speech recognition, natural language processing, audio recognition, bioinformatics, and medical science It is no exception for the SAR remote sensing community. Lin et al [21] in 2017; and the anchor-free CenterNet proposed by Duan et al [22] in 2019
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have