Learning-based multi-view stereo has made significant advances. However, some methods neglect the guidance of various geometric priors, which leads to poor reconstruction under challenges like occlusions and weak textures. Meanwhile, other works only explore the contribution of a single geometric prior at a relatively fixed stage, which lacks holistic constraints and complementary collaboration. Therefore, this paper integrates surface normals, epipolar geometry, visibility, and geometric consistency through the entire pipeline. It aims to resolve various issues through the collaboration of multiple complementary geometric cues at different stages. The Key innovations include a dual guidance of surface normals in feature extraction and volume aggregation, which enhances the network’s robustness to occlusion through geometric visibility and epipolar-enhanced feature group correlation, while also reinforcing geometric structural features. Along with a frequency-optimized consistency inpainting pipeline through multi-direction and multi-stage. With the further utilization of depth optimization modules, extensive experiments on 5 datasets of varying scales demonstrate that our proposed MGPG-MVSNet exhibits excellent generalization to real-world engineering applications, particularly achieving superior reconstruction in occluded and textureless regions.

Building instance segmentation has a wide range of applications in remote sensing images, but faces the challenges of preserving geometric details and adapting to multi-scale objects in complex backgrounds. To address these problems, this paper proposes a new Directional Fields-Guided Building Instance Segmentation Network with Unpooling Content Reconstruction (URDGNet). Specifically, URDGNet comprises an Unpooling Content Reconstruction (UCR) module and a Background-Percepted Directional Fields Guidance (BP-DG) module. The UCR achieves the content reconstruction through an Unpooling Content Attention strategy and a Multi-level Content Compensation strategy. In particular, unpooling technology is used to restore spatial dimensions and expand channel weights, while the receptive field is extended to compensate for content details across different feature scales. In addition, the BP-DG achieves more accurate edge extraction and background suppression by synergistically exploiting global context modeling and local directional field constraints. Experiments on the WHU, Vaihingen, and CTCities datasets show that URDGNet improves the mean average precision (mAP) by 0.9% to 9.0% compared to eight state-of-the-art instance segmentation methods, and by 1.4% to 6.1% compared to ten state-of-the-art object detection methods, validating its effectiveness and adaptability.

Cross-view geo-localization has been widely used as an important technique for determining the geographical location of unmanned aerial vehicles (UAV). Despite various image retrieval methods proposed, drone and satellite image cross-view geo-localization still remains challenging due to their wildly inconsistent view angles. In this paper, we propose a new framework, Swin-Radial-Locality Network (SRLN), to extract robust image feature representations. Specifically, SRLN is based on a pruned version of the Swin Transformer, which integrates multi-scale feature aggregation within a Siamese network structure, featuring shared weights and equipped with multi-classification heads. SRLN is mainly comprised of a Radial-Slicer-Network (RSN) and a Local-Pattern-Network (LPN), which is designed to effectively harmonize directional information from drone-captured images and broader environmental features from satellite imagery, crucial for capturing angle and feature details between drone and satellite images. The RSN part focuses on capturing fine-grained features that represent the drone’s directional information, while the LPN is utilized for a more comprehensive analysis of broader environmental features. Extensive experiments are carried out on widely used public benchmark datasets, i.e., University-1652 and SUES-200. With more than 3% improvement over existing methods in both drone-view target localization tasks and drone navigation applications, the results validate the superior performance of our multi-scale feature fusion model, achieving a state-of-the-art performance record.

Cross-view image matching aims to match images of the same target scene acquired from different platforms. With the rapid development of drone technology, cross-view matching by neural network models has been a widely accepted choice for drone position or navigation. However, existing public datasets do not include images obtained by drones at different heights, and the types of scenes are relatively homogeneous, which yields issues in assessing a model’s capability to adapt to complex and changing scenes. In this end, we present a new cross-view dataset called SUES-200 to address these issues. SUES-200 contains 24120 images acquired by the drone at four different heights and corresponding satellite view images of the same target scene. To the best of our knowledge, SUES-200 is the first public dataset that considers the differences generated in aerial photography captured by drones flying at different heights. In addition, we developed an evaluation for efficient training, testing and evaluation of cross-view matching models, under which we comprehensively analyze the performance of nine architectures. Then, we propose a robust baseline model for use with SUES-200. Experimental results show that SUES-200 can help the model to learn highly discriminative features of the height of the drone.

As the energy consumption of multi-core systems becomes increasingly prominent, it's a challenge to design an energy-efficient real-time scheduling algorithm in multi-core systems for reducing the system energy consumption while guaranteeing the feasibility of real-time tasks. In this paper, we focus on multi-core processors, with the global Dynamic Voltage Frequency Scaling (DVFS) and Dynamic Power Management (DPM) technologies. In this setting, we propose an energy-efficient real-time scheduling algorithm, the Time Local remaining execution plane based Dynamic Voltage Frequency Scaling (TL-DVFS). TL-DVFS utilizes the concept of Time Local remaining execution (TL) plane to dynamically scale the voltage and frequency of a processor at the initial time of each TL plane as well as at the release time of a sporadic task in each TL plane. Consequently, TL-DVFS can obtain a reasonable tradeoff between the real-time constraint and the energy-saving while realizing the optimal feasibility of sporadic tasks. Mathematical analysis and extensive simulations demonstrate that TL-DVFS always saves more energy than existing algorithms, especially in the case of high workloads, and guarantees the optimal feasibility of sporadic tasks at the same time.