Few-shot segmentation (FSS) aims to segment objects of interest in a query image using a limited set of support images. However, most existing FSS methods are designed for natural images. When extended to remote sensing scenes characterized by extreme intra-class variations and complex backgrounds, these methods struggle to provide robust segmentation guidance, leading to severe performance degradation. To address the aforementioned issues, we propose a multimodal visual-language prompt network (MVLPNet), which employs a collaborative optimization strategy for visual-textual features to tackle the remote sensing FSS task. Specifically, MVLPNet consists of a textual-visual consistency enhancement (TVCE) module and a prototype-guided semantic alignment (PGSA) module. To overcome the limited support set for better guiding the query segmentation, we propose a TVCE module that leverages the contrastive language-image pre-training model (CLIP) to capture category-specific text embeddings. An optimal transport (OT) plan is then established to tightly align these text embeddings with the visual features of query image, thereby extracting semantic information from the query image itself to mitigate the extreme intra-class variation in remote sensing images. Furthermore, a PGSA module is proposed to suppress interference caused by complex background regions. By aggregating lost foreground regions, more comprehensive support features are extracted. Then, the query and support features are precisely matched to activate consistent foreground regions, rather than ambiguously matching the query features via a single prototype or multiple prototypes. Extensive experiments on the iSAID-5i and LoveDA-2i datasets have demonstrated that our method achieves the state of the art. The code is available https://github.com/Gritiii/MVLPNet.

Few-shot domain adaptive object detection (FSDAOD) aims to transfer knowledge from a source domain to a target domain with limited labeled data, which faces severe challenges in the field of remote sensing. To address this issue, numerous CNN-based domain adaptation methods employ style transfer and feature alignment to mitigate domain shifts, but limited target domain samples are prone to yielding equivocal optimization and are maladaptive. Furthermore, the sparsity of targets and the complexity of backgrounds in remote sensing imagery contribute to confusing feature alignment. Moreover, DETR-based detectors have achieved remarkable progress in unsupervised domain adaptation (UDA) but remain unexplored in FSDAOD. To address these challenges, we introduce FSDA-DETR, the first DETR-based strong baseline designed for the FSDAOD of remote sensing imagery. Specifically, we propose a cross-domain style rectification (CSR) module that rectifies the styles of the target domain to align with the source domain by storing and dynamically updating the weighted-fusion source domain style bases. To further strengthen the detector’s cross-domain detection performance, we propose a category-aware feature alignment (CFA) module that performs fine-grained masking on object regions of rectified backbone features corresponding to different categories and utilizes adversarial training for domain-invariant feature extraction. Extensive experiments on three cross-domain benchmarks, comprising six diverse datasets, demonstrate that FSDA-DETR outperforms state-of-the-art methods. For instance, in the Optical-to-SAR benchmark, FSDA-DETR achieves 74.5% mAP with only 1% of target domain training data. The code and datasets are available at https://github.com/wsybb252237/FSDA-DETR.

Tiny object detection represents a pivotal challenge in remote sensing intelligent interpretation, necessitating detectors to exhibit heightened precision in object localization. However, typical model optimization strategies cannot release the detector’s potential for precisely localizing objects. And the lack of interpretability in detection box filtering based on object classification scores serves as a constraint on further performance improvement. Therefore, this paper proposed a novel model optimization strategy to thoroughly unleash the potential of the detector for precise localization. Then, the utilization of object comprehensive confidence score enhances the interpretability of the post-processing step for detection boxes. Rigorous experiments on the AI-TOD dataset have demonstrated the effectiveness of our method, achieving state-of-the-art performance.

Remote sensing images often present complex backgrounds, introducing significant noise that hinders feature extraction and object detection, particularly for small objects. This paper introduces the Multilevel Feature-Aware Hybrid Encoder (MFA-Encoder), designed to efficiently extract and fuse features by selectively employing different attention modules at various levels within the encoder. In this hybrid encoder, we leverage self-attention, cross-channel cross-attention, and multiscale compound attention instead of applying self-attention mechanisms to features of all scales. Expand the dimensions of high-level features, reinforce the representation of low-level features. Comprehensive experiments demonstrate that our model achieves best performance with an AP of 65.5% on the DIOR dataset.

Few-shot object detection (FSOD) from optical remote sensing imagery has to detect rare objects given only a few annotated bounding boxes. The limited training data is hard to represent the data distribution of realistic remote sensing scenes, restricting the performance of FSOD. Recently, learning conditional controls for text-to-image diffusion model has achieved great progress, which is capable of precisely generating the controllable yet imaginational images by text prompt and spatially localized input conditions. Accordingly, in this work, we aim to explore the potential of diffusion model and propose a solution for few-shot object detection by controllable data generation. Firstly, draw upon a few annotated objects, their bounding boxes and categories are respectively used as the spatial conditions and text prompts, then employ them into large text-to-image diffusion models for controlled image generation. Secondly, based the generated images, in order to adapt to the scale and orientation variances of remote sensing objects, a data transformation is devised for boosting the robustness of model training. Finally, some experiments were conducted on public remote sensing dataset DIOR, and the results proved its effectiveness.

Abstract Remote sensing target tracking in satellite videos plays a key role in various fields. However, due to the complex backgrounds of satellite video sequences and many rotation changes of highly dynamic targets, typical target tracking methods for natural scenes cannot be used directly for such tasks, and their robustness and accuracy are difficult to guarantee. To address these problems, an algorithm is proposed for remote sensing target tracking in satellite videos based on a variable‐angle‐adaptive Siamese network (VAASN). Specifically, the method is based on the fully convolutional Siamese network (Siamese‐FC). First, for the feature extraction stage, to reduce the impact of complex backgrounds, we present a new multifrequency feature representation method and introduce the octave convolution (OctConv) into the AlexNet architecture to adapt to the new feature representation. Then, for the tracking stage, to adapt to changes in target rotation, a variable‐angle‐adaptive module that uses a fast text detector with a single deep neural network (TextBoxes++) is introduced to extract angle information from the template frame and detection frames and performs angle consistency update operations on the detection frames. Finally, qualitative and quantitative experiments using satellite datasets show that the proposed method can improve tracking accuracy while achieving high efficiency.

Remote sensing target tracking in UAV (Unmanned Aerial Vehicle) aerial videos has gradually become a research hotspot of UAV application areas, in this paper, we propose a remote sensing target tracking method for UAV aerial videos based on multi-frequency feature enhancement. The main contributions are as follows: first, We use SiamRPN as the basic tracking network to obtain the natural ability of multi-scale testing, which can adapt to the scale changing of remote sensing targets caused by changes in shooting perspective. Second, in the stage of feature extraction, we introduce a new multi-frequency feature representation method, which effectively improves feature expression ability of highly dynamic targets. Our algorithm is compared quantitatively and qualitatively with some state-of-the-art tracking algorithms by using a test dataset consisting of a UAV123 dataset and a homemade dataset, the results show that our algorithm improves the tracking accuracy while having strong timeliness.

Airport aircraft detection is a research hotspot in the field of automatic target detection in optical remote sensing images. The existing detection methods generally have low efficiency and poor detection accuracy due to the uncertainty of aircraft scale and orientation and to the complexity of airport remote sensing images. To address these problems, this paper presents an effective aircraft detection framework called Local Context DPM (LC-DPM). Our method is conducted in two main stages. (1) During aircraft candidate region extraction, we propose a non-flat region extraction method and a regular region elimination method to extract aircraft candidate regions in airport. (2) We propose LC-DPM during identification of the aircraft candidate region, we achieve accurate identification by constructing the local context histogram of oriented gradients (HOG) feature pyramids, which combine the local context information and HOG features of the aircraft targets. In addition, in order to further improve the efficiency of the method, we use the circle- frequency filter to predict the orientation of the suspected aircraft target, before the identification at various orientations based on LC-DPM. We test the proposed method on complex airport area sets with varying types of aircraft. The experimental results show that the proposed method is both highly accurate and computationally efficient.

Inshore ship detection from high-resolution (HR) optical satellite images is a hot research field. However, HR ships multi-scale and multi-orientation characters and harbor scene various interferences affect docked ship detection performance. Therefore, we proposed a multi-orientations sparse dictionaries (MOSDs) algorithm combining with comprehensive structure voting (CSV) to address existed problem and achieve refined docked ship contour region proposal (RP). Moreover, the comparing experiments use a lot of Google Earth harbour images to demonstrate proposed method effectiveness and robustness of HR ships multi-scale and -orientation changing and various harbour background interferences of docked ship detection.

With the development of optical remote sensing satellite, ship detection and identification from large-scale remote sensing images has become a priority research topic. Specially, inshore ship detection has received increasing attention in many safe and marine applications. However, most of the popular techniques for inshore ship detection are limited by calculation efficiency and detection accuracy. In this paper, for inshore ship detection in complex harbor areas, we present a novel hierarchical method combining an efficient candidate scanning and a cascade model strategy. First, in the phase of candidate regions extraction, we design an omnidirectional intersected two-dimension (OITD) scanning method to extract candidate regions from the land-water segmented images rapidly. In addition, in candidate region identification phase, we structure a cascade model strategy to identify real ships from candidates to improve the accuracy of identification. The cascade model strategy is integrated by a bow model and a hull model of ship, which are trained by Deformable Part Model (DPM). Experiments on large-scale harbor remote sensing images show the higher precision and rapid computational efficiency of the proposed method.

The middle wave infrared remote (MWIR) images has complex scene information, low contrast ratios, and bipolar problems. To solve these problems, we propose a method that uses pyramid integral image reconstruction algorithm achieving sea-land automation segmentation. First, we calculate a gradient feature map (GFM), which extracts the structural information from an MWIR scene. Then, the GFM uses for sum are table (SAT) generation. The pyramid integral image reconstruction technology uses different scale factor reconstruct MWIR images by using SAT. Then the adaptive threshold method is employed for the sea-land segmentation on the multi-scale integral reconstruction images. Finally, we get sea-land refine segmentation result of MWIR images by synthesis analysis the multi-scale reconstruction images. By using GFM and pyramid integral image reconstruction operation, are avoid with the complex gray scene information. The integral image reconstruction can enhance the structure information and improve the reconstruction image contrast. This paper proposed method is from structure and texture information view point for sea-land segmentation, so the bipolar problem is solve in our method for MWIR images sea-land segmentation.

Synthetic aperture radar (SAR) is an indispensable and extensively used sensor in ship detection. As high-resolution SAR introduces more spatial details into images, this letter proposes an intensity-space (IS) domain constant false alarm rate (CFAR) ship detector to make good use of this information. The method fuses intensity of each pixel and correlations between pixels into one characteristic, i.e., IS index. All the detection procedures center on the calculation and analysis of IS index. First, a new transform maps an image into a new IS domain. Structures like ships and wakes are enhanced in IS domain. Second, a CFAR detector picks up high IS index pixels. Third, a chain of target features is checked to screen out false candidate target pixels. Also, enhanced wakes are taken to improve detection results. Experiments on real SAR images validate that the proposed transform does enhance these structures and the whole algorithm is of good performance, especially in the case of low-contrast targets.

Airplane detection is a challenging problem in complex remote sensing imaging. In this letter, an effective airplane detection framework called Markov random field-fully convolutional network (M-FCN) is proposed. The M-FCN uses a cascade strategy that consists of an FCN-based coarse candidate extraction stage, a multi-Markov random field (multi-MRF)-based region proposal (RP) generation stage, and a final classification stage. In the first stage, the FCN model is trained to be sensitive to airplanes, and a coarse candidate map is generated. This model is scale-, direction-, and color-invariant and does not require many training examples. After the first stage, the coarse candidate map is used as the initial labeling field for a multi-MRF algorithm, and RPs are generated according to the multi-MRF output. This RP-generating strategy can yield more accurate locations with fewer RPs. In the last stage, a convolutional neural network-based classifier is used to improve the precision of the entire framework. Experiments show that the M-FCN has high precision, recall, and location accuracy.

Harbor water area extraction is a key step in nearshore environment pollution surveillance using remote sensing image processing techniques. This letter proposes the definition circle (DC) model of color gradient to describe color fluctuations in harbor water surface areas based on pan-sharpened remote sensing images. The DC model includes two steps: center setting and radius tuning. In the center setting process, labeled training set pixels are selected in the red, green, and blue color space. Then, center setting is completed in the hue, saturation, and intensity color space using the perceptron model. In the radius tuning process, positive and negative sample pixels are used to tune the radius value. After these two steps, the DC model can describe the color gradient of a water surface area and provide accurate harbor water area extraction. A series of experiments shows that the proposed DC model is robust and performs better than other extraction methods based on pan-sharpened remote sensing images.

The measurement of optical fibre vibration is a key part of optic fibre pre‐warning system, which has gradually focused on phase‐sensitive optical time‐domain reflectometer. However, for this instrument, false alarm rate is very high and some unstable intrusion signals cannot be detected by using its fixed threshold method in the actual application. It needs to develop new vibration detection method to overcome the above defect. The vibration signals normally consist of three parts, that is, noise, interference and intrusion signals. After a large number of data analysis, the authors find that the system noise is time varying and follows the Rayleigh distribution. Hence, the authors innovatively use the constant false alarm rate (CFAR) method to detect this type of intrusion. Considering interference is also time varying and diverse, a good detection performance cannot be obtained only by using the conventional CFAR. For this reason, a background homogeneity adaptive CFAR (BHA‐CFAR) method is further proposed to detect the vibration signals in this study. The BHA‐CFAR consists of two detectors, cell averaging CFAR (CA‐CFAR) detector and greatest‐of/smallest‐of CFAR (GO/SO‐CFAR) detector. A parameter, homogeneity of background, is estimated first to classify the surrounding. Then CA‐CFAR and GO/SO‐CFAR are optionally used according to the surrounding is homogeneous or heterogeneous, respectively. This new detection method can adapt to any background surrounding and has a good detection performance. In order to check the feasibility and validity of the BHA‐CFAR method, several experiments were carried out in Da Gang oilfield. The detection results show that the proposed method can provide a good tradeoff between the detection performance and computation time.

This letter proposes a synthetic aperture radar (SAR) image registration method named feature-area optimization (FAO). First, the traditional area-based optimization model is reconstructed and decomposed into three key but uncertain factors: initialization, slice set, and regularization. Next, structural features are extracted by scale-invariant feature transform (SIFT) in dual-resolution space (SIFT-DRS), a novel SIFT-like method dedicated to FAO. Then, the three key factors are determined based on these features. Finally, solving the factor-determined optimization model can get the registration result. A series of experiments demonstrate that the proposed method can register multitemporal SAR images accurately and efficiently.

Automatic urban area detection in remote sensing images is an important application in the field of earth observation. Most of the existing methods employ feature classifiers and thereby contain a data training process. Moreover, some methods cannot detect urban areas in complex scenes accurately. This letter proposes an automatic urban area detection method that uses multiple features that have different resolutions. First, a downsampled low-resolution image is used to segment the candidate area. After the corner points of the urban area are extracted, a weighted Gaussian voting matrix technique is employed to integrate the corner points into the candidate area. Then, the edge features and homogeneous region are extracted by using the original high-resolution image. Using these results as the input, the processes of guided filtering and contrast enhancement can finally detect accurately the urban areas. This method combines multiple features, such as corner, edge, and regional characteristics, to detect the urban areas. The experimental results show that the proposed method has better detection accuracy for urban areas than the existing algorithms.

In this letter, we propose a novel computational model for automatic ship detection in optical satellite images. The model first selects salient candidate regions across entire detection scene by using a bottom-up visual attention mechanism. Then, two complementary types of top-down cues are employed to discriminate the selected ship candidates. Specifically, in addition to the detailed appearance analysis of candidates, a neighborhood similarity-based method is further exploited to characterize their local context interactions. Furthermore, the framework of our model is designed in a multiscale and hierarchical manner which provides a plausible approximation to a visual search process and reasonably distributes the computational resources. Experiments over panchromatic SPOT5 data prove the effectiveness and computational efficiency of the proposed model.