Impact of Geometric Priors: Advanced Fine-grained Airplane Detection with Geometric Details in High-resolution Satellite Images

Evaluating Classical and Deep Keypoint Detectors for SfM Reconstruction in Arctic UAV Imagery

Abstract. Structure-from-Motion (SfM) pipelines rely heavily on the detection and matching of repeatable keypoints across images, yet the performance of modern learned feature extractors in challenging environments remains insufficiently understood. This paper evaluates classical and deep keypoint detectors for SfM reconstruction using winter Arctic UAV imagery, a domain characterized by low texture, repetitive patterns, and limited man-made structure. We compare three feature pipelines within a shared PyCOLMAP-based framework: SIFT with nearest-neighbor matching (SIFT+NN), SuperPoint, and DISK, along with a hybrid approach combining SuperPoint and DISK correspondences. Quantitative evaluation is conducted using standard SfM metrics, including number of observations, track length, observations per image, and reprojection error, complemented by qualitative analysis of keypoint distributions and reconstruction interpretability. Results show that SIFT+NN consistently achieves the most complete and stable reconstructions, producing the highest number of matched observations and lowest reprojection error across aggregate experiments. However, on more challenging subsets lacking clear structural features, learned methods demonstrate improved robustness, successfully reconstructing multiple views where SIFT fails. SuperPoint provides broader spatial coverage, while DISK produces denser clusters in high-confidence regions, highlighting complementary behaviors between learned approaches. Overall, the findings indicate that classical methods remain strong baselines for Arctic UAV photogrammetry under standard SfM pipelines, while learned detectors offer advantages in difficult conditions. The observed performance gap is attributed to domain mismatch and backend optimization for handcrafted features. These results suggest that domain-specific training and improved spatial feature distribution are promising directions for advancing learned keypoint methods in Arctic reconstruction tasks.

Bundle-Adjusted Initialization for Efficient Earth Observation Gaussian Splatting

Abstract. Satellite imagery offers a distinct advantage in Earth observation by providing expansive coverage and enabling the monitoring of inaccessible regions without physical on-site intervention, serving as a significantly more cost-effective and scalable alternative to traditional aerial or ground-based surveys. The task of 3D reconstruction from multi-view satellite images has therefore been a pivotal point of research at the intersection of photogrammetry and remote sensing. Recently, novel-view synthesis techniques such as Neural Radiance Fields (NeRF) and 3D Gaussian Splatting (3DGS) have accelerated the accuracy and speed of topographic modeling. Among these, Earth Observation Gaussian Splatting (EOGS) has emerged as a state-of-the-art approach by adapting 3DGS to handle the unique geometric and radiometric characteristics of satellite data, including Rational Polynomial Coefficients (RPCs) and varying solar conditions. However, the standard EOGS pipeline relies on stochastic initialization, where Gaussians are distributed uniformly within a volumetric bounding box, leading to high computational overhead and dependency on aggressive pruning that can inadvertently remove critical geometric features, particularly in areas with complex urban structures. To address these limitations, we propose Bundle-Adjusted Initialization for Earth Observation Gaussian Splatting, which leverages sparse point clouds from bundle adjustment as geometric priors for Gaussian initialization. Combined with an adaptive densification strategy, our method achieves faster convergence and improved DSM accuracy on the DFC2019 dataset compared to the EOGS baseline.

Geometry-Consistent Content-Aware Fusion for Lightweight Small-Object Detection in UAV Imagery

Small-object detection in unmanned aerial vehicle (UAV) imagery is challenging because targets are often tiny, densely distributed, and embedded in cluttered backgrounds, while edge platforms impose strict computational constraints. This paper presents a lightweight end-to-end detector based on RT-DETR to improve detection accuracy, localisation quality, and deployment efficiency. The proposed method includes three components: a detail-enhanced backbone for preserving fine-grained texture and edge cues, rotary positional encoding combined with a content-aware bidirectional feature pyramid for improved cross-scale alignment, and a difficulty-aware composite regression loss for more stable bounding-box optimisation. Experiments on VisDrone2019 and DOTA show that the proposed method improves mAP50 by 3.5% and 1.6%, respectively, over the RT-DETR baseline, while reducing the number of parameters by 28.6%. Ablation studies confirm the complementary contributions of the three components, and robustness experiments demonstrate reliable performance under challenging aerial conditions. Deployment on a Jetson Orin NX using a self-built UAV dataset further validates the practicality of the method for edge-side UAV monitoring.

Beyond Appearance: Camouflaged Object Detection via Geometric Structure

A Frequency-Aware and Multiscale Aligned Framework for Fine-Grained Object Detection in Remote Sensing Imagery

Fine-grained object detection in remote sensing imagery is challenged by small targets, subtle interclass differences, and cross-scale feature mismatch. This article proposes a frequency-aware and multiscale aligned framework for YOLO-based detectors. FAENet separates and enhances low-frequency structural information and high-frequency details before backbone feature extraction. DSAF aligns adjacent pyramid features through joint pooling and interpolation to reduce scale mismatch in the neck. ScalSeq with ASF Attention aggregates multiscale features and refines the $P_{3}$ branch to strengthen small-target discrimination. Experiments on MAR20, HRSC2016, and ShipRSImageNet cover four YOLO families, three model scales, and three input resolutions. The full configuration improves detection accuracy in most settings and generalizes across aircraft and ship targets. Among the evaluated full configurations, YOLOv5-M achieves the best accuracy–cost balance: its average accuracy increases from 0.524 to 0.550, while the parameter count rises from 25.076 to 26.023 M and inference time from 1.4 to 2.2 ms, yielding the highest accuracy–cost gain ratio of 0.087. Comparisons with representative non-YOLO detectors further confirm its practical efficiency.

Semantic Consistency and Uncertainty-Driven Small-Object Detection for Class Imbalance

In aerial image small-object detection, complex imaging perspectives, arbitrary object orientations, and long-tailed category distributions jointly exacerbate sample imbalance, which significantly degrades detection stability and leads to frequent misclassification of minority categories. To address these challenges, this paper proposes a novel training framework termed SCUD. Specifically, in the label noise suppression strategy (LNSS), a contrastive learning mechanism based on semantic consistency is introduced to constrain the aggregation of similar samples in the feature space, thereby reducing the adverse impact of noisy samples on model optimization. In addition, a scale-aware resampling strategy (SARS) is designed to alleviate noise amplification and overfitting caused by excessive repetition of small objects during training. Furthermore, an adaptive instance selection mechanism (AISM) is developed by jointly modeling prediction uncertainty and global statistical priors, enabling the model to dynamically emphasize learning from informative samples. Extensive experiments are conducted on two publicly available unmanned aerial vehicle (UAV) aerial image datasets to validate the effectiveness of the proposed approach. The proposed method achieves an mAP50 of 70.7% on the DOTA-v1.0 dataset and 88.1% on the DIOR dataset. Notably, the detection accuracy of several rare categories is significantly improved, further demonstrating the effectiveness of the proposed method in addressing sample imbalance in aerial image small-object detection.