The rapid development of remote sensing technology has led to a sharp increase in the amount of synthetic aperture radar (SAR) measurements, which put forward higher requirements for remote sensing image processing. As an important application of SAR, fast and accurate ship detection has always been a research hotspot. In this article, an improved lightweight RetinaNet for ship detection in SAR images is proposed. Compared with the standard RetinaNet, the shallow convolutional layers of the backbone are replaced by ghost modules and the number of the deep convolutional layers is reduced. The spatial and channel attention modules are embedded into the model to enhance detectability. K -means clustering algorithm is applied to adjust the initial aspect ratios of the model. The effectiveness and robustness of the proposed method is demonstrated by numerical experiments with SSDD dataset, Gaofen-3 mini dataset, and a large-scale SAR image of Hisea-1 satellite, it is shown that the proposed method can significantly reduce the floating-point operations and the number of parameters without decreasing the detection accuracy and recall ratio. Moreover, the experimental results also show the proposed model's robustness and the ability to detect ship targets in small datasets.

Ocean surveillance via high-resolution Synthetic Aperture Radar (SAR) imageries has been a hot issue because SAR is able to work in all-day and all-weather conditions. The launch of Chinese Gaofen-3 (GF-3) satellite has provided a large number of SAR imageries, making it possible to marine targets monitoring. However, it is difficult for traditional methods to extract effective features to classify and detect different types of marine targets in SAR images. This paper proposes a convolutional neutral network (CNN) model for marine target classification at patch level and an overall scheme for marine target detection in large-scale SAR images. First, eight types of marine targets in GF-3 SAR images are labelled based on feature analysis, building the datasets for further experiments. As for the classification task at patch level, a novel CNN model with six convolutional layers, three pooling layers, and two fully connected layers has been designed. With respect to the detection part, a Single Shot Multi-box Detector with a multi-resolution input (MR-SSD) is developed, which can extract more features at different resolution versions. In order to detect different targets in large-scale SAR images, a whole workflow including sea-land segmentation, cropping with overlapping, detection with MR-SSD model, coordinates mapping, and predicted boxes consolidation is developed. Experiments based on the GF-3 dataset demonstrate the merits of the proposed methods for marine target classification and detection.

Polarimetric synthetic aperture radar (PolSAR) image semantic segmentation is currently of great importance for synthetic aperture radar (SAR) image interpretation, especially in agricultural applications. Several convolutional neural networks (CNNs) have been implemented for SAR image semantic segmentation in urban or land cover applications. However, existing CNNs often break one semantic area into several pieces or confuse the adjacent semantic area in a PolSAR image. To address these issues uniformly, a refined pyramid scene parsing network (PSPNet) is proposed for PolSAR image semantic segmentation in an agricultural areas. Compared to conventional PSPNet architecture, the refined PSPNet adopts a multilevel feature fusion design in its decoder to effectively exploit the features learned from its different encoder branches. Besides, a polarimetric channel attention module is incorporated into the network to capture rich polarimetric features in a PolSAR image. Furthermore, an edge-aware loss function is devised to guide the network to refine pixel-level edge information directly from semantic segmentation prediction, separating easily confused agriculture area with sharp contours. Experimental results on one airborne millimeter-wave PolSAR dataset verify that the proposed network achieves promising semantic segmentation accuracy and preferable spatial consistency.

Recent years have witnessed rapid development and remarkable achievements on deep learning object detection in remote sensing (RS) images. The growing improvement of the accuracy is inseparable from the increasingly complex deep convolutional neural network and the huge amount of sample data. However, the under-fitting neural network will damage the detection performance facing the difficulty of sample acquisition. Thus, it evolves into few-shot object detection (FSOD). In this article, we first briefly introduce the object detection task and its algorithms, to better understand the basic detection frameworks followed by FSOD. Then, FSOD design methods in RS images for three important aspects, such as sample, model, and learning strategy, are respectively discussed. In addition, some valuable research results of FSOD in computer vision field are also included. We advocate a wide research technique route, and some advice about feature enhancement and multi-modal fusion, semantics extraction and cross-domain mapping, fine-tune and meta-learning strategies, and so on, are provided. Based on our stated research route, a novel few-shot detector that focuses on contextual information is proposed. At the end of the paper, we summarize accuracy performance on experimental datasets to illustrate the achievements and shortcomings of the stated algorithms, and highlight the future opportunities and challenges of FSOD in RS image interpretation, in the hope of providing insights into future research.

Global Navigation Satellite System (GNSS)-based bistatic Synthetic Aperture Radar (SAR) recently plays a more and more significant role in remote sensing applications for its low-cost and real-time global coverage capability. In this paper, a general imaging formation algorithm was proposed for accurately and efficiently focusing GNSS-based bistatic SAR data, which avoids the interpolation processing in traditional back projection algorithms (BPAs). A two-dimensional point target spectrum model was firstly presented, and the bulk range cell migration correction (RCMC) was consequently derived for reducing range cell migration (RCM) and coarse focusing. As the bulk RCMC seriously changes the range history of the radar signal, a modified and much more efficient hybrid correlation operation was introduced for compensating residual phase errors. Simulation results were presented based on a general geometric topology with non-parallel trajectories and unequal velocities for both transmitter and receiver platforms, showing a satisfactory performance by the proposed method.

Long time coherent integration is a vital method for improving the detection ability of global navigation satellite system (GNSS)-based passive radar, because the GNSS signal is not radar-designed and its power level is very low. For aircraft detection, the large range cell migration (RCM) and Doppler frequency migration (DFM) will seriously affect the coherent processing of azimuth signals, and the traditional range match filter will also be mismatched due to the Doppler-intolerant characteristic of GNSS signals. Accordingly, the energy loss of 2-dimensional (2-D) coherent processing is inevitable in traditional methods. In this paper, a novel 2-D coherent integration processing and algorithm for aircraft target detection is proposed. For azimuth processing, a modified Radon Fourier Transform (RFT) with range-walk removal and Doppler rate estimation is performed. In respect to range compression, a modified matched filter with a shifting Doppler is applied. As a result, the signal will be accurately focused in the range-Doppler domain, and a sufficiently high SNR can be obtained for aircraft detection with a moving target detector. Numerical simulations demonstrate that the range-Doppler parameters of an aircraft target can be obtained, and the position and velocity of the aircraft can be estimated accurately by multiple observation geometries due to abundant GNSS resources. The experimental results also illustrate that the blind Doppler sidelobe is suppressed effectively and the proposed algorithm has a good performance even in the presence of Doppler ambiguity.