Change detection (CD) in optical remote sensing images has significantly benefited from the development of deep convolutional neural networks (CNNs) due to their strong capability of local modeling in bi-temporal images. In addition, the recent rise of transformer modules leads to the improvement of global feature extraction of bi-temporal remote sensing images. Note that the existing simple cascade of deep CNNs and transformer modules shows limited CD performance on small changed areas due to deficiencies of multi-scale information therein. To address the aforementioned issue, we propose a new CNN-transformer network (ConvTransNet) with multi-scale framework to better exploit global-local information in optical remote sensing images. In our ConvTransNet, we propose the parallel-branch ConvTrans block as the basic component to generate global-local features, i.e., adaptively integrates the global features summarized by a transformer-based branch and the local features extracted by a convolution-based branch, providing better identifiability between changed areas and unchanged areas. By fusing multiple global-local features with different scales, our ConvTransNet improves the robustness of the CD performance on changed areas with different sizes, especially small changed areas. Experiments on two public change detection datasets of optical remote sensing images, i.e., LEVIR-CD and CDD, demonstrate that our ConvTransNet achieves enhanced CD performance than the other commonly used methods.

Convolutional neural networks (CNNs) have been widely applied in the context of ship detection in synthetic aperture radar (SAR) images, but the detection performance is still not ideal in scenarios with clutter interference. Mining frequency-domain information to suppress the sea clutter in SAR ship detection has attracted wide attention. However, existing frequency-domain ship detection methods do not process frequency-domain information adaptively, which results in the degradation of ship detection performance. To overcome this problem, this article proposes a novel deep learning network called YOLO-FA. YOLO-FA contains the proposed Frequency Attention Module (FAM), which can process frequency-domain information of SAR images adaptively. The proposed method can suppress the sea clutter in the SAR images with the help of frequency-domain information. We evaluate the proposed method YOLO-FA on two datasets, i.e., the High-Resolution SAR images dataset (HRSID) and SAR Ship Detection Dataset (SSDD). Compared with the baseline method YOLOv5 and existing commonly used methods, YOLO-FA achieves state-of-the-art detection performance on both datasets.

Existing superpixel-based detection algorithms for ship targets in synthetic aperture radar (SAR) images are often derived from the local contrast of intensities (i.e., the local contrast of the first-order information of superpixels) leading to deteriorating performance in low signal-to-clutter ratio (SCR) cases due to the low contrast between the intensities of targets and the clutter. In this article, we propose a new superpixel-based detector to improve the performance of ship target detection in SAR images via the local contrast of fisher vectors (LCFVs). The new LCFV-based detector exploits multiorder features of the superpixels based on the Gaussian mixture model (GMM) and accordingly improves the discrimination capability between the ship targets and the sea clutter, especially in low SCR cases. Experimental results demonstrate that the proposed LCFV-based detection algorithm provides better detection performance than the commonly used detection algorithms.

Image matching is a primary technology for optical and synthetic aperture radar (SAR) image fusion but often shows limited performance due to the highly nonlinear differences between optical and SAR modalities. Recently, deep neural networks (DNNs) have been investigated to effectively extract nonlinear features for image matching tasks, where DNNs are trained based on the elaborated design of loss functions and a low loss value is often expected to obtain better image matching performance. In this letter, we first theoretically demonstrate that when the value of a state-of-the-art loss function decreases, the corresponding matching performance may not consistently improve due to the imbalanced effect of positive and negative samples. To tackle this issue, we proposed an improved loss function to train DNNs for image matching of SAR and optical images. We theoretically prove that the improved loss function ensures the improvement of the matching performance when the loss value decreases based on Taylor’s series expansion analysis. Experimental results on an open dataset with extensive optical and SAR image pairs show that 1) the proposed loss function is better than the original one in terms of image matching performance and 2) the combination of our loss function and existing multiscale convolutional gradient feature (MCGF)-based network provides better matching performance than other state-of-art approaches.

In this letter, we propose a new constant false alarm rate (CFAR) detector to accelerate the existing superpixel (SP)-based CFAR detectors for ship detection in synthetic aperture radar (SAR) images. In our method, we design a new density-censoring operation to rapidly identify background clutter SPs (BCSPs) with high densities before the local CFAR detection. In this way, a large number of non-informative BCSPs are removed without time-consuming calculation of decision thresholds, and only a few candidate ship target SPs (STSPs) are retained. This reduces the computational cost of the subsequent local CFAR detection and the number of false alarms produced by it. During the local CFAR detection process for the retained candidate STSPs, we also propose an improved method to define their neighboring clutter regions (for the calculation of decision thresholds) using BCSPs identified by the density-censoring operation. Experiments on measured SAR images validate that the proposed CFAR method reduces the computational cost of commonly used SP-based CFAR methods by 75%-96% with similar or better detection performance.

Recently, convolution neural networks (CNNs) have been extensively utilized in synthetic aperture radar (SAR) ship detection owing to their strong feature extraction and representation capability. However, existing CNN-based SAR ship detectors often suffer from poor sensitivity to small-scale ship targets due to the limited extractable features, especially in complex inshore scenarios. Moreover, the hand-designed components like non-maximum suppression calculation and anchor generation imposed in CNN-based detector significantly affect their robustness. In face of the these challenges, a novel end-to-end transformer-based detection framework for small-scale ship targets in SAR images, named DINO (DETR with Improved deNoising anchOr box) with Refined De-noising and Box (RDB-DINO), is proposed in this paper. Firstly, we introduce a complete contrastive de-nosing (CCD) training technique which reconstructs and exploits different kinds of noised queries to reduce the confusion between small ships and complex background. Secondly, a look twice towards maximum (LTTM) algorithm for iterative box refinement is designed to mine the abnormal sample information and obtain abundant features of small ships in training process. Finally, substantial experiments conducted on two widely-used open SAR ship datasets demonstrate that the proposed approach yields superior results in small ship detection performance, outperforming prevailing state-of-the-art benchmarks.