Accurate satellite tracking requires up-to-date Two-Line Elements (TLEs), as outdated data can lead to significant positioning errors. While ground-based optical telescopes are highly accessible, generating TLEs from their data is complicated by the fundamental lack of direct range measurements. This paper presents an automated end-to-end pipeline designed to overcome this limitation by proposing a robust method to estimate the range from prior TLE. The pipeline can then generate the updated TLEs by calculating new satellite state vectors using the estimated range. The pipeline consists of star and satellite detection, astrometric calibration, and orbit determination. For star and satellite detection, the key component of the pipeline is a robust deep learning-based detection model. To achieve this, we benchmarked models such as Deformable DETR, RF-DETR, and YOLOv12 against traditional image processing methods with 3105 images in FITS (Flexible Image Transport System) format. RF-DETR yields the highest F1 score (0.93) and precision (0.97) at an 8-pixel threshold. The detected star coordinates resulting from using RF-DETR, the best-performing model, were fed into Astrometry.net for precise astrometric calibration to determine the satellite celestial coordinates. The range required for orbit determination was estimated by extracting the prior TLE and propagating it to the observation epoch via SGP4. The satellite state vectors were then calculated using TLE-constrained orbit determination approach using the estimated range, followed by an inverse SGP4 optimization to recover the mean orbital elements. The generated TLEs were validated against public TLEs from Space-Track.org. To evaluate this pipeline, updated TLEs were generated specifically for medium Earth orbit (MEO) and geostationary Earth orbit (GEO) targets. The results demonstrate that the proposed method yields high accuracy for GEO satellites, achieving a mean motion difference of 0.0030 rev/day and a 24-hour ground-track position error of 0.83 degrees. In comparison, MEO satellites achieve a mean motion difference of 0.0126 rev/day and an error of 5.27 degrees. These results suggest that the proposed pipeline provides a robust foundation for automated orbit determination with clear potential for further refinement.
Abstract. In recent decades, many different satellites for Earth observation have been launched. They produce large amounts of data that, if properly preprocessed, can be used in many applications. A rapidly growing portion of these data comes from small satellites, which remain underused by scientists and entrepreneurs. However, greater utilisation can be ensured by producing images with positional accuracy of at least two pixels, which is necessary for their reliable use. This paper presents the upgraded version of the geometric correction module of the STORM processing chain. The module can automatically orthorectify images from the NEMO-HD small satellite, which, like other small satellites, in principle has a lower signal-to-noise ratio (SNR) and higher radiometric variability. It automatically extracts ground control points (GCPs) by matching freely available reference vector roads and reference images to roads extracted from the satellite image using a deep learning method trained on PlanetScope and NEMO-HD imagery. The performance of the geometric correction module was evaluated using three images acquired over Slovenia. The road extraction method can achieve an F1-score of approximately 60%. The tests demonstrated that automatic GCP extraction based on roads detected by a deep learning method is a viable approach for achieving geometric model accuracies of two pixels or less at independent check points when using small satellite imagery.
Aleš Marsetič, P. Pehani, Nina Krašovec· The International Archives o...· 0 citations
High-resolution spacecraft images provide important astrometric constraints for orbit refinement, but measurements of resolved bodies are often limited by labor-intensive control-point selection and the difficulty of achieving consistent reductions over large image archives. We present an automated shape-model-based astrometric pipeline for Phobos and apply it to Mars Express Super Resolution Channel (SRC) images. For each exposure, a synthetic image is rendered from a high-resolution 3D shape model under the nominal spacecraft-target-Sun geometry. Feature correspondences between the observed and synthetic images are established using SuperPoint and SuperGlue, followed by RANSAC filtering. The matched synthetic-image keypoints are then associated with surface points through ray-shape intersection. The geometric adjustment fixes the adopted body orientation, spacecraft state, and corrected camera pointing and estimates only two effective plane-of-sky position offsets using the exact perspective-projection model. These offsets are used to derive the center-of-figure position of Phobos. We first test the method on an image set previously analysed with a control-point approach and obtain comparable astrometric performance. We then extend the analysis to a larger SRC dataset spanning 2007-2025 and obtain 1113 successful measurements. Relative to the JPL MAR099 ephemeris, the resulting observed-minus-computed residuals have mean values of 0.186 km in $\alpha \times cos(\delta)$ and 0.053 km in $\delta$, with corresponding standard deviations of 0.609 km and 0.583 km. These results demonstrate that the proposed pipeline provides a practical approach to large-scale, homogeneous astrometric reduction of archival spacecraft images of Phobos, with potential application to other resolved bodies.
Wangxin Lai, Qingfeng Zhang, Rui Zhang et al.· 0 citations
Of nearly 1.5 million asteroids discovered to date, only about 5% have spectroscopic observations, resulting in limited physical property coverage. Multicolor photometry provides taxonomic classifications for the majority of asteroids. Meanwhile, asteroids originating from the same collisional event and occupying nearby regions of orbital space typically share similar surface properties. Consequently, orbital parameters can provide statistical constraints on asteroid type and albedo, enabling more reliable estimates of their sizes. To achieve accurate inversion of asteroid size and albedo using readily available information under optical observation and to improve the completeness of asteroid physical property databases, this paper proposes a deep learning method that takes an 11-dimensional vector comprising absolute magnitude, orbital parameters, and type information as input, and embeds the Bowell formula to fuse physical priors with data-driven modeling. The proposed method eliminates the dependence on spectral data, thereby significantly expanding the training sample size, which in turn ensures robust model generalization. Test results on asteroids explored by spacecraft or radar observations show mean absolute percentage errors of 26.4% and 21.6% for albedo and effective diameter, respectively. By applying the model to all asteroids with known types, a large-scale asteroid physical properties catalog was constructed. The resulting catalog contains taxonomic type, geometric albedo, and effective diameter for 188,737 asteroids. Compared to existing catalogs based on machine learning predictions, the sample coverage has improved by an order of magnitude. The generated catalog is available on Zenodo (doi:10.5281/zenodo.20032293).
Jiayi Ge, Xiaoming Zhang, Juan Li et al.· Astrophysical Journal Supple...· 0 citations
Modern astronomical surveys generate large volumes of data and provide precise astrometric and photometric reference catalogues. Efficient use of these datasets requires automated, robust, and instrument-independent reduction pipelines.
We introduce (Photometry and Astrometry of Point Sources), a fully automated Python-based pipeline for photometric reduction with integrated astrometric calibration. It supports the homogeneous reduction of stellar and moving Solar System objects, independent of telescope aperture or detector type. PhoPS
PhoPS performs astrometric calibration by dynamically generating local Gaia Data Release 3 (DR3) reference index files propagated to the epoch of observation, avoiding the need for a pre-installed index collection. For photometry, it adopts a field-dependent calibration strategy in which the zero point (ZP) is modelled across the detector plane using Random Sample Consensus (RANSAC)-based linear regression, thereby accounting for spatial systematics such as vignetting and detector non-uniformities.
The astrometric contribution was evaluated by comparing solutions obtained with epoch-propagated and non-propagated local Gaia DR3 index files. Using 141,672 matched measurements from 840 images obtained with the 1-m TUG100 telescope, the propagated solution improves the clipped N-weighted total root-mean-square (RMS) residual from 0.284 to 0.241 a 15.0% improvement. The photometric uncertainty model was validated using 203 reference stars and 92,980 measurements. The normalised residuals are centred close to zero. However, the residual width is magnitude dependent: the limited bright-star bin (10 łeq G < 13) shows excess scatter with σ_z=2.069, the intermediate bin (13 łeq G < 16) is closest to unit variance with σ_z=1.129, and the faintest bin (16 łeq G < 18) is conservative with σ_z=0.698. Thus, the validation reveals a magnitude-dependent uncertainty behaviour rather than a globally consistent error model.
PhoPS provides an open-source, lightweight, cross-platform solution for astronomical image reduction. It is suited to asteroid light-curve analysis and stellar variability studies and can diagnose systematics related to telescope tracking, focus stability, and vignetting.
Orhan Erece, Yücel Kılıç· Astronomy & Astrophysics· 1 citation
Processing the lunar astrophotography imagery is challenging since atmospheric turbulence and low-light conditions introduce blur and noise that
obscure small-scale lunar features. The present work proposes an enhancement pipeline that operates entirely within classical signal processing,
providing a transparent alternative to black-box machine learning methods while remaining practical on standard personal computers. The pipeline
integrates four key stages: lucky-imaging frame selection and stacking, wavelet-domain denoising, adaptive local contrast enhancement, and
dual-stage edge sharpening. The pipeline is evaluated on 30 lunar datasets spanning multiple phases and multiple observable conditions. Quantitative
results show Peak Signal-to-Noise Ratio (PSNR) improvements of approximately 6.2–12.4 dB and an increase in Structural Similarity Index
Measure (SSIM) from 0.56 to 0.90, indicating better structural fidelity relative to stacked baselines. Stacking of 16 carefully selected frames yields
effective SNR gains of up to about 3.2 times, while modulation transfer function (MTF) analysis at limb and crater-edge boundaries reveals sharpness
improvements in the order of 28–35%. The workflow requires no specialized accelerators, with typical resource usage of roughly 4.2 MB memory
and 2.3 seconds per megapixel on conventional CPUs. This is demonstrating that classical, interpretable techniques remain highly competitive for
scientific and educational lunar image enhancement.
S. Bhattacharya· Asian journal of applied sci...· 0 citations
With 50,000 current actively tracked objects in orbit and a clear upwards trajectory, especially as launch vehicle technologies improve, orbital density and the amount of assets in space is increasing, making space debris a more and more pressing issue. As active debris removal is difficult to make sustainable, active prevention through precise satellite tracking becomes necessary. While current ground based solutions are limited in update rate and are incredibly expensive making them hard to scale, this paper investigates the feasibility of a decentralized alternative where data gathered from a global network of amateur astronomers, classified using a machine learning model that automatically detects satellite trails in telescope imagery, providing position updates at higher frequency and lower costs than conventional systems. A ResNet-50 convolutional neural network was used for trail detection, trained on more than 11,000 images and achieving an overall accuracy of 98% and F1 score of 0.95 on a validation set demonstrating feasibility of use with a full network. A separate global network simulation using TLE orbital data modelling variations of a global network over a 48 hour period found that performance scaled almost linearly, with the best configuration of 50 stations globally and 5 observers at each station achieving an update rate more than 11x higher than the U.S. Space Force’s Space Surveillance Network baseline at a fraction of the cost. The results demonstrate both technical and economic viability of the decentralized network, with the potential to play a very meaningful role in preserving sustainable access to Earth orbit.
Jay Zhang· International Journal of Ene...· 0 citations