Recently, ViTAE-RVSA, the first large-scale Vision Transformer (ViT) tailored for remote sensing, has demonstrated the potential of ViTs by integrating window attention with a convolutional branch. However, as this branch still adopts isotropic square kernels and conventional Feature Pyramid Networks neglect directional priors in multi-scale fusion, the resulting representations remain suboptimal for arbitrarily oriented, high-aspect-ratio, and small objects. To overcome these limitations, we propose Strip-ViTAE, a transformer-based detector that embeds a rotation-equivariant convolution (R2Conv) in the backbone to capture orientation-consistent local features. Moreover, we introduce a Parallel Strip Convolution Module (PSCM) employing horizontal and vertical strip kernels to adaptively enlarge the receptive field and strengthen local feature modeling for elongated and small objects. Finally, we design a Strip Feature Pyramid Network (StripFPN) composed of a Bottom-Up Reflow Module (Bottom-Up RM), a Strip-Based Enhancement Module (SBEM), and a Self-Attention Excitation Module (SAEM) to achieve direction-consistent cross-scale fusion. Experiments on the DOTA-v1.5 dataset show that Strip-ViTAE attains 72.82% mAP, surpassing the state-of-the-art (SOTA) by 1.35% and improving the mAP for high-aspect-ratio and small objects by 3.20%, validating the effectiveness of our method.

Fine-grained aircraft classification in optical remote sensing images relies on subtle difference of local parts. However, aircrafts in low-resolution (LR) images are depicted as small objects with blurred details and limited appearance information, resulting in the dramatic classification performance drop. We propose a novel dynamic hybrid knowledge distillation (DHKD) method to enhance fine-grained aircraft classification in LR remote sensing images. The core idea is to construct a teacher network for high-resolution (HR) images and a student network for LR images, and dynamically distill the key discriminative region attention-guided feature-based and logit-based object prior knowledge acquired from the teacher to the student to instruct its training, thereby improving the student’s fine-grained representation capability and classification performance for LR images. The distillation weights of different knowledge types are adaptive to their optimization direction similarity with that of the primary fine-grained classification task during the iterations. Extensive experiments on the MAR20 dataset demonstrate that DHKD consistently outperforms state-of-the-art distillation methods on fine-grained classification for aircrafts with absolute sizes of 32, 20, and 12 pixels in LR remote sensing images.

Recent deep learning based-works have made remarkable progress in Remote-Sensing Image Super-Resolution (RSISR). However, the complicated network architecture as well as a huge amount of parameters increase computational cost, hindering their practical deployment. To alleviate this problem, we propose a novel Re-parameterized Feature Distillation Network (ReFDN) for lightweight and efficient RSISR tasks. Feature distillation, refinement, condensation, and enhancement are efficiently integrated into the re-parameterized feature distillation block named ReFDB for lighter and stronger feature extraction. With the help of elaborate re-parameterized convolution (ReConv) design, we further boost the feature refinement capability without extra inference costs. Additionally, we design an efficient channel and spatial attention module (ECSA) to enhance the important objects and regions of the intermediate features adaptively. Conducted on both commonly used datasets and additional Google Earth data, the experimental results demonstrate our method can achieve a good trade-off between SR performance and network complexity. Our code will be publicly available at https://github.com/DaxingZ/ReFDN.

Fine-grained ship classification (FGSCR) has many applications in military and civilian fields. In recent years, deep learning has been widely used for classification tasks, and its success is inseparable from that of big data. However, ship images are valuable, with only a few images of a specific category being obtained, leading to the fine-grained few-shot ship classification problem. In addition, feature map channels contain distinct characteristics and discriminative details, which significantly influence FGSCR. Intuitively, channels with distinct characteristics should be assigned larger weights for classification, but most few-shot learning methods treat the channels equally. Therefore, we propose a generalized ridge-regression-based channelwise feature map weighted reconstruction network to address these issues. First, we reconstruct the query feature map by assigning different weights to the support feature map channels using the generalized ridge regression method. The channels with large discriminative details contribute more toward reconstruction. Second, we propose a support channel weight module to calculate the channel weight matrix used in the generalized ridge regression method. Finally, based on the reconstructed query feature map, we can calculate the reconstruction error. The reconstruction error is adopted as the distance metric. Our proposed method achieves excellent performance on the fine-grained ship, bird, aircraft, and WHU-RS19 datasets compared with other representative few-shot learning methods. Considering the limited studies on the fine-grained few-shot ship classification problem, we believe that our work is of great significance.

Interpreting airborne remote sensing images plays an important role in aviation control and battlefield situational awareness. However, highly dynamic aircraft detection remains challenging, owing to variable object sizes, flexible attitudes, and motion blur. This study develops a multi-scale airborne aircraft dataset benchmark to overcome aircraft detection challenges, such as high intraclass variance, multiple scales and angles, motion blur, and partial occlusion. We also propose a trained-from-scratch aircraft detector, the local feature correlation single shot multibox detector (LFC-SSD), to detect multi-scale aircraft. The LFC-SSD comprises a local correlation feature extraction module, called “Right-Residual,” and a feature fusion module using a reverse feature pyramid network (R-FPN). Right-Residual extends the global receptive field by aggregating contextual information while learning non-adjacent region features efficiently. R-FPN utilizes multi-path information transfer horizontally with recursive integration to enhance the robust representation of the location information of multi-scale object features. In addition, a specific default boxes method is designed for remote-sensing images of aircraft. Extensive experimental results confirm the significant improvement of the proposed method over several existing state-of-the-art methods.

Object tracking in satellite videos faces various challenges such as target occlusion, target rotation, and background clutter. This study proposes a correlation particle filter algorithm with motion estimation for object tracking in satellite videos. The tracker, called CPKF, combines the strengths of the correlation, particle, and Kalman filters. Compared with existing tracking methods based on correlation filters, the proposed tracker has three major advantages: (1) Particle sampling, and motion estimation build robustness against partial and complete occlusion. (2) Color histogram model makes it robust to target rotation. (3) Fusion of multiple feature response maps effectively handle background clutter and low contrast. The experimental results demonstrate that the proposed tracking algorithm performs better than state-of-the-art methods.

In synthetic aperture radar (SAR) images, the edge of the shadow is blurred because the radar is moving while the data are collected. In this letter, this problem is expanded on by using the imaging formation perspective. First, an approximate method to represent the imaging quality of the boundary of the shadow region based on the quadratic phase errors (QPEs) is provided for the first time, which built up the relationship between the parameters of the shadow caster and the behavior of the shadow in the SAR image. We notice that the QPE is approximately a linear function of the height of the caster. Second, we deduced the height-dependent phases due to the synthetic aperture process to the raw data, and a novel algorithm called height-variant phase compensation (HVPC) on the complex SAR image data is proposed by compensating the unexpected phases in the azimuth to sharpen the shadow. Compared with the traditional approach called fixed-focus shadow enhancement (FFSE), HVPC removes twice as much of the QPE as FFSE approximately. Experiments on simulation and real data demonstrate the precision and the better effect on shadow enhancement of our work. It is expected that the work in this letter could be some help for the SAR image understanding and application.

In this letter, a stationary-wavelet-based despeckling algorithm based on the two-sided generalized gamma distribution (GΓD) model is proposed. We first introduce the two-sided GΓD as a flexible and efficient model for the wavelet coefficients of logarithmically transformed synthetic aperture radar intensity or amplitude. The strength of the model is highlighted in terms of its fit to the data, its low computational cost, and the ease of parameter estimation. By empirical results, we then motivate the GΓD as model for the wavelet coefficients of the noise-free signal. The GΓD model parameters are estimated with moment methods, using both absolute central moments for the wavelet coefficients of the noisy signal and the noise. Finally, we exploit the prior information contained in the model by designing a Bayesian maximum a posteriori estimator for estimating the noise-free wavelet coefficients. Experimental results demonstrate the superiority of our method in terms of simultaneously reducing speckle and preserving structural details.

In this letter, a novel synthetic aperture radar (SAR) imaging simulator is proposed based on analytical electromagnetic and geometric models for urban structures. The backscatter contributions are evaluated by an analytical electromagnetic model based on the Kirchhoff approach (KA) in either physics or geometrical optics approximations rather than specular and Lambertian models. In addition, the position vectors of object facets in the SAR imaging plane are evaluated by a closed-form analytical geometrical model based on the ray-tracing model. These models are expressed in terms of few and basic parameters. Compared with other numerical methods, they are helpful for improving the efficiency of the simulation, but more importantly, they are significant for direct understanding of and further interpreting the SAR image features. Some experiments validate the geometric and electromagnetic models and demonstrate the efficiency of the simulator.