The rapidly growing number of space activities is generating numerous space debris, which greatly threatens the safety of space operations. Therefore, space-based space debris surveillance is crucial for the early avoidance of spacecraft emergencies. With the progress in computer vision technology, space debris detection using optical sensors has become a promising solution. However, detecting space debris at far ranges is challenging due to its limited imaging size and unknown movement characteristics. In this paper, we propose a space debris saliency detection algorithm called SDebrisNet. The algorithm utilizes a convolutional neural network (CNN) to take into account both spatial and temporal data from sequential video images, which aim to assist in detecting small and moving space debris. Firstly, taking into account the limited resource of the space-based computational platform, a MobileNet-based space debris feature extraction structure was constructed to make the overall model more lightweight. In particular, an enhanced spatial feature module is introduced to strengthen the spatial details of small objects. Secondly, based on attention mechanisms, a constrained self-attention (CSA) module is applied to learn the spatiotemporal data from the sequential images. Finally, a space debris dataset was constructed for algorithm evaluation. The experimental results demonstrate that the method proposed in this paper is robust for detecting moving space debris with a low signal-to-noise ratio in the video. Compared to the NODAMI method, SDebrisNet shows improvements of 3.5% and 1.7% in terms of detection probability and the false alarm rate, respectively.

The increasing amount of space debris in recent years has greatly threatened space operation. In order to ensure the safety level of spacecraft, space debris perception via on-orbit visual sensors has become a promising solution. However, the perception capability of visual sensors largely depends on illumination, which tends to be insufficient in dark environments. Since the images captured by visible and infrared sensors are highly complementary in dark environments, a convolutional sparse representation-based visible and infrared image fusion algorithm is proposed in this paper to expand the applicability of visual sensors. In particular, the local contrast measure is applied to obtain the refined weight map for fusing the base layers, which is more robust in a dark space environment. The algorithm can settle two significant problems in space debris surveillance, namely, improving the signal-noise ratio in a noise space environment and preserving more detailed information in a dark space environment. A space debris dataset containing registered visible and infrared images has been purposely created and used for algorithm evaluation. Experimental results demonstrate that the proposed method in this paper is effective for enhancing image qualities and can achieve favorable effects compared to other state-of-the-art algorithms.

Accurate recognition and tracking of circular markers constitutes the cornerstone of high-fidelity dynamic displacement measurement in videogrammetry. However, conventional trackers fail in Structural Health Monitoring (SHM) due to three mechanically induced conflicts: (1) the visual homogeneity of identical markers invalidates appearance-based Re-Identification (Re-ID), causing identity switches; (2) nonlinear motions, such as rotation, violate the linear motion assumptions of standard Kalman filters, leading to prediction divergence; and (3) prolonged physical occlusions sever temporal continuity, making trajectory recovery impossible for motion-blind algorithms. To address these issues, we propose Circle-Tracker, a geometry-constrained stereo tracking system that exploits the inherent spatial rigidity of marker groups. To resolve visual ambiguity, the framework transforms the complex Multi-Object Tracking (MOT) problem into parallel Single-Object Tracking (SOT) tasks via a Dynamic Region-of-Interest (ROI) strategy, which enforces local spatial consistency. To overcome nonlinear motion and occlusion, a dedicated Geometry-Aware Recovery Module is introduced. This module leverages group kinematics to rectify predictions and integrates a novel homography pipeline (AGCR and RBMS) to robustly re-establish identity by mapping historical topology to current observations. Extensive validation against a proprietary dataset demonstrates that Circle-Tracker significantly outperforms mainstream baselines, exhibiting superior robustness in Re-ID accuracy.

Super resolution (SR) of digital elevation models (DEMs) is crucial for numerous geospatial applications, yet existing methods still face severe challenges when dealing with multi-scale reconstruction and multi-source data. In particular, mainstream DEM SR approaches primarily rely on convolutional neural networks (CNNs), whose inherent local receptive fields limit their ability to model the global structural consistency of large geomorphic units such as mountain ranges. As a result, when handling terrains with spatial heterogeneity, it is difficult to simultaneously ensure the fidelity of local details. To address this feature representation bottleneck, this paper proposes a unified framework named TAAF-SRNet. The framework integrates collaborative learning with meta-learning-based upsampling to enable flexible multi-scale reconstruction. At its core are the enhanced feature extraction module (EFEM), which adopts a parallel CNN-Transformer architecture, and a novel terrain aware adaptive fusion mechanism. This mechanism can dynamically assess local terrain complexity and adaptively adjust the weighting between the local textures captured by CNNs and the global context modeled by Transformers. The terrain aware strategy enables a unified model to adapt to diverse landforms, ranging from vast abyssal plains to rugged terrestrial mountain ranges. Extensive experiments on different DEM datasets covering three representative heterogeneous geomorphologies, namely land, land-sea transition, and seabed, demonstrate that TAAFSRNet not only outperforms current state-of-the-art (SOTA) methods on image quality metrics such as PSNR and LPIPS, but also better preserves terrain fidelity, as reflected by lower MAE and RMSE in elevation, slope, and aspect estimation. The code will be available at: https://github.com/bbqboy666/TAAF-SRNet.