In complex environments, traditional multi-target tracking methods often encounter challenges such as strong clutter interference and interruptions in target trajectories, which can result in insufficient tracking accuracy and robustness. To address these issues, this paper presents an improved multi-target tracking algorithm, termed Q-IMM-MHT. This method integrates Multiple Hypothesis Tracking (MHT) with Interactive Multiple Model (IMM) and introduces a Q-learning-based adaptive model switching strategy to dynamically adjust model selection in response to variations in the target’s motion patterns. Furthermore, the algorithm utilizes Support Vector Machines (SVMs) for anomaly detection and trajectory recovery, thereby enhancing the accuracy of data association and the overall robustness of the system. Experimental results indicate that under high noise conditions, the Root Mean Square Error (RMSE) of position estimation decreases to 0.74 pixels, while the RMSE of velocity estimation falls to 0.04 pixels/frame. Compared to traditional methods such as the Unscented Kalman Filter (UKF), IMM, and CIMM, the RMSE is reduced by at least 10.84% and 42.86%, respectively. In scenarios characterized by target trajectory interruptions and clutter interference, the algorithm maintains an association accuracy exceeding 46.3% even after 30 frames of interruption, significantly outperforming other methods. These findings demonstrate that the Q-IMM-MHT algorithm offers substantial performance improvements in multi-target tracking tasks within complex environments, effectively enhancing both tracking accuracy and stability, with considerable application value and extensive potential for future use.

The automatic, accurate perception of targets in space is a crucial prerequisite for many on-orbit aerospace missions. Therefore, research on perception technologies within spaceborne images is meaningful. The development of deep learning has revealed its potential for application to space target perception. However, implementing deep learning models requires large-scale labelled datasets. Therefore, we build a multitask synthetic benchmark space target dataset, NCSTP, to address the limitations of current datasets. First, we collect and modify various space target models for satellites, space debris, and space rocks. By importing them into a realistic space environment simulated by Blender, 200,000 images are generated with different target sizes, poses, lighting conditions, and backgrounds. Then, the data are annotated to ensure the dataset supports simultaneous space target detection, recognition and component segmentation. All data can be used for training space target detection and recognition models. We further annotate the components of each satellite for component segmentation. Finally, we test a series of state-of-the-art object detection and semantic segmentation models on the dataset to establish a benchmark.

Multispectral remote sensing image super-resolution (RSISR) aims to reconstruct high-frequency details while preserving cross-band structural consistency under strict computational budgets. However, real-world satellite imagery exhibits heterogeneous distortions, ranging from band-dependent noise to spatially varying texture degradation, rendering uniform restoration strategies suboptimal. To address these challenges, we propose a unified framework that integrates cue extraction, expert specialization, and efficiency-aware restoration. Specifically, a Distortion-Aware Feature Extractor (DAFE) explicitly encodes distortion cues by synthesizing fixed frequency bases, learnable residual components, lightweight spatial edge representations, and noise proxies. Subsequently, a Distortion-Aware Expert Choice (DAEC) router utilizes these cues to establish distortion-conditioned affinities and performs capacity-constrained, load-balanced expert assignment. Finally, a parameter-shared Mixture-of-Experts (PS-MoE) architecture employs shared expert parameters across spectral bands, augmented by band-wise low-rank adapters, to enable coarse-to-fine restoration with minimal computational overhead. Extensive experiments on the SEN2VENμS and OLI2MSI datasets demonstrate that the proposed method achieves a PSNR of 49.38 dB on SEN2VENμS 2×, 45.91 dB on SEN2VENμS 4×, and 45.94 dB on OLI2MSI 3×. Compared to the strongest baseline for each task, our method yields PSNR improvements of 0.12 dB, 0.10 dB, and 0.09 dB, respectively, while simultaneously reducing FLOPs and parameter counts. These results confirm that explicit distortion modeling and parameter-shared expert specialization provide an effective and computationally efficient solution for multispectral remote sensing image super-resolution.

Unresolved or nearly unresolved infrared remote sensing small targets captured by lightweight imaging systems are often severely degraded by field-dependent blur and noise, leading to energy dispersion, reduced local visibility, and degraded interpretability. This work proposes a physics-informed spatially variant image restoration method for this target regime. Unlike conventional restoration methods based on a spatially invariant degradation assumption, the proposed framework explicitly incorporates a spatially variant point spread function (PSF) prior into a learning-based restoration pipeline to model field-dependent imaging degradation. On this basis, a deep Wiener-style restoration mechanism is introduced to jointly recover target energy and suppress noise, while a matched-filter-inspired PSF-correlation-guided soft attention module is designed to emphasize target likelihood regions and mitigate clutter amplification during restoration. In addition, a target-oriented optimization strategy is adopted to promote local target observability for weak small targets under complex infrared backgrounds. By combining physical imaging priors with adaptive deep restoration, the proposed method offers an interpretable solution for spatially variant infrared small-target recovery when accurate local PSFs are available. On a controlled synthetic benchmark with calibrated PSF priors, the proposed method achieves a mean local signal-to-noise ratio (SNR) gain of 12.56 dB.