Reference-based super-resolution (Ref-SR) is a heated topic distinguished from single-image super-resolution (SISR). It aims at transferring more texture details from reference (Ref) image with a different perspective, to super-resolve the low-resolution (LR) image. However, the development of Ref-SR within remote sensing (RS) community is limited by three problems. First, RS images exhibit more complex texture details compared to natural images. It’s challenging to learn and reconstruct fine texture of RS images. Second, the lack of high-quality RS image dataset, which contains massive RS image pairs from different perspectives, hampers the model training and diminishes the generalization of Ref-SR within RS community. Third, the lack of physical system prevents it from verifying the feasibility of Ref-SR in RS practice. To address these problems, this paper proposes a novel reference-based gradient-assisted texture-enhancement GAN (RGTGAN), a novel dataset, namely KaggleSRD, and a novel physical simulation system, namely dual-zoom-lens system (DZLS). Specifically, this paper proposes a gradient-assisted texture-enhancement module (GTEM) to fully release the potential of gradient branch to learn fine structures during feature extraction process, a novel dense-intern deformable convolution (DIDConv) to boost the alignment effect between features from different image branches during feature alignment process, and a novel dense-restore-residual (DRR) module to effectively transfer features. Extensive experimental results on both datasets, RRSSRD and KaggleSRD, demonstrate the superiority of the proposed method over state-of-the-art methods. Furthermore, DZLS verifies promising application prospects of the proposed method. Our code and dataset are publicly available at: https://github.com/stdinR/RGTGAN.

In the previous study, there were a few direct star identification (star-ID) algorithms for smearing star image. An end-to-end star-ID algorithm is proposed in this article, to directly identify the smearing image from star sensors with fast attitude maneuvering. Combined with convolutional neural networks and the self-attention mechanism of transformer encoder, the algorithm can effectively classify the smearing image and identify the star. Through feature extraction and position encoding, neural networks learn the position of stars to generate semantic information and realize the end-to-end identification for the smearing star image. The algorithm can also solve the problem of low identification rate due to smearing of long exposure time for images. A dataset of dynamic stars is analyzed and constructed based on multiple angular velocities. Experiment results show that, compared with representative algorithms, the identification rate of the proposed algorithm is improved at high angular velocities. When the three-axis angular velocity is 10°/s, the rate is still 60.4%. At the same time, the proposed algorithm has good robustness to position noise and magnitude noise.

Polarization and infrared imaging technology have unique advantages for various applications ranging from biology to ocean remote sensing. However, conventional combined polarization camera and infrared camera have limitations because they are constrained to single-band imaging systems with rotating polarizers and cascaded optics. Therefore, we propose a common-aperture mode based on multi-band infrared guided polarization imaging system (IGPIS) in this paper, which consists of infrared wide-area sensing and polarization features acquisition for accurate detection of ship targets. The IGPIS can provide images in visible polarization (0.45–0.76 μm), near-infrared polarization (0.76–0.9 μm), and long-wave infrared (8–12 μm) bands. Satellite attitude parameters and camera optical parameters are accurately calculated by establishing a dynamic imaging model for guidance imaging. We illustrate the imaging principle, sensors specifications and imaging performance analysis and the experimental results show that the MTF is 0.24 for visible and near-infrared, and 0.13 for long-wave infrared. The obtained multi-band images have an average gradient of 12.77 after accurate fusion. These results provide theoretical guidance for the design of common-aperture cameras in remote sensing imaging field.

Compared with traditional push-broom imaging, the space camera that performs the new circular scanning imaging can effectively enlarge the imaging area and obtain high-resolution images. However, the imaging velocity fields of space camera will seriously deteriorate the image quality in the circular scanning imaging process. And the velocity fields are anisotropically distributed due to various factors, including attitude maneuvering, satellite precession, earth curvature, and earth rotation. Therefore, this paper proposes a mathematical model of the velocity field to analyze the effect of circular scanning imaging. Based on the principle of ray tracing, the expression of the instantaneous velocity field on the ground is first derived, and then the corresponding image velocity field is obtained by mapping the velocity vector to the focal plane. In addition, numerical simulations and image simulations are performed to analyze the anisotropy distribution of the velocity vectors. Moreover, the dynamic characteristics of the space camera are analyzed to evaluate the performance of this new imaging. The simulation results show that the higher velocity leads to worse imaging quality. The study of this paper provides guidance for attitude planning and imaging assessment for new remote sensing imaging of space cameras.

When the targets are distributed irregularly, it is of great significance to quickly acquire color images with a wide field by agile satellite. Attributed to the changing of the three-axis attitude and Earth’s rotation, there are misalignment and distortion during the processing of imaging by agile satellite. It makes the fusion of interband images slower and the stitching of interdetector images poorer. In this article, a rigorous velocity vector field (VVF) model is established through tracing rays between the image and the object. It calculates the misalignment among the images of bands in each detector to align and fuse multispectral images. Based on the reference image selected by VVF, the fused images are stitched with the method of leveraging line-point consistency. Finally, both field and in-orbit experiments are conducted. The time of calculating misalignment in image fusion with VVF is 13.19% of the minimum time of the traditional registration algorithm. In image stitching, our RMSE values are 1.15, which is the smallest among the RMSE values of SIFT, SURF, and SPW. In preserving linear structure, compared with SPW, the values of ${E_{\text {dis}}}$ , ${E_{\text {err}}}$ , and ${E_{\text {cross}}}$ , respectively, reduce 60%, 50%, and 50%. The results show that images of the agile satellite can be fast fused and precisely stitched using VVF while correcting distortion. The compensation strategy of the VVF provides a theoretical basis for the agile satellite in image fusion and stitching.

Dynamic analysis of the optical–mechanical structure based on the linear state space expression is performed for a three-point flexural mount lightweight space mirror with a diameter of ϕ740 mm. Using linear structure dynamics and linear state space theories, the state space model of the mirror assembly is established based on modal information. The DC gain method is used to reduce modes and a frequency response analysis of the reduced modes is performed to obtain the frequency domain transfer function between the excitation input and response output points and determine the contribution of each mode to the total frequency response. The frequency response curve is plotted. A mechanical vibration test is performed to verify the accuracy and rationality of simulation analysis. The dynamic analysis method based on state space theory provides a new method of investigating optical and mechanical structures, which can help efficiently and accurately analyze the frequency response characteristics of complex linear systems.