Visual-inertial odometry (VIO) serves as a dominant framework for real-time motion state estimation in micro aerial vehicles (MAVs). However, existing VIO techniques remain highly susceptible to high dynamic range (HDR) illumination conditions. Bio-inspired by the lateralized visual system of the Strawberry Squid, we propose a SSS-VIO technique that integrates a functionally asymmetric-eye design and perceptual attention mechanisms to overcome HDR-induced limitations. The proposed method is applicable both to the design of asymmetric stereo platforms and to cost-effective retrofitting of commercial stereo cameras. Experimental evaluations on an OAK-4P-New camera demonstrate a 40% increase in usable dynamic range with a single neutral density filter. Furthermore, SSS-VIO also outperforms multiple state-of-the-art VIO methods in MAV flight tests, reducing Absolute Trajectory Error by at least 46.1% and consistently improving Relative Pose Error across different trajectory segments. To the best of our knowledge, this is the first bio-inspired asymmetric-eye solution for enhancing the VIO dynamic range, which might unlock new potentials for MAV indoor-outdoor localization and applications.

Conventional noise estimation methods for arrayed microelectromechanical system (MEMS) inertial measurement units (MIMUs) exhibit reduced effectiveness in dynamic scenarios, leading to degraded measurement accuracy and stability. To address this problem, this article proposes an adaptive robust fusion method for the MIMU array designed for dynamic applications. The performance enhancement mechanism of the MIMU array is analyzed, and the theoretical basis of the reduced effectiveness of traditional noise estimation methods under dynamic conditions is investigated. A second-order redundant observation mechanism is introduced to enable reliable online noise estimation of the MIMU array in dynamic environments. Building upon this estimation, a fading-memory factor and a similarity evaluation strategy are incorporated to achieve adaptive and robust fusion of multiple MIMU measurements. Simulation results demonstrate that the proposed method can effectively estimate the noise level of the MIMU array in dynamic scenarios, while vehicle experiments further verify its capability to enhance both accuracy and robustness in practical dynamic applications.

In the field of satellite navigation and positioning, the selection of satellites is crucial for enhancing positioning accuracy, reducing computational load, and optimizing resource utilization. With the operation of the four global navigation satellite systems, the number of visible satellites has increased dramatically. Under the influence of multipath errors and non-line-of-sight reception, the quality of satellite signals becomes uneven in complex environments. Selecting the optimal set of satellites from a large number of visible ones with varying quality, in order to achieve the best balance between positioning accuracy and real-time performance, is a challenging task. To address this issue, the paper conducts a detailed analysis of the factors influencing positioning accuracy. It develops an evaluation model for positioning accuracy contributions that incorporates both satellite observation quality and geometric configuration, referred to as the AWGDOP (Adaptive Weighted Geometric Dilution of Precision) contribution. This evaluation model allows for the assessment of each satellite's contribution to the overall positioning accuracy. Furthermore, a fast satellite selection algorithm, based on positioning accuracy contribution, is proposed. It achieves rapid satellite selection by iteratively eliminating the satellite with the smallest contribution. To verify the effectiveness of the fast satellite selection algorithm proposed in this paper, tests are conducted in both open and obstructed environments. The test results indicate that the proposed algorithm can reduce the positioning error by 25% to 35% in terms of accuracy and decrease computational time by 5% to 20% in terms of real-time performance, compared to the full-constellation solution based on an empirical model.

The Fried parameter and coherence radius are two fundamental measures that characterize the spatial resolution effects of atmospheric turbulence on optical propagation and imaging. Here, their calculations for spherical and plane waves in von Karman turbulence are theoretical developed that includes the effects of both nonzero inner scale and finite outer scale. The simple analytic expressions are also derived for making the results easy to use. Then the split-step wave optics simulations are performed for modeling the long-exposure point spread function and mutual coherence function through von Karman turbulence for comparison. The expressions and numerical results agree well throughout the weak to strong turbulent scattering regimes.

Multirobot collaborative SLAM, as a key direction in Internet of Things (IoT) applications, faces the critical challenge of achieving accurate self-pose and mutual-pose estimation. Existing distributed pose optimization methods often suffer from unbalanced computational allocation and insufficient utilization of inter-robot information. To address these limitations, we propose a balanced and robust distributed pose optimization (BR-DPO) framework that integrates three complementary layers: inter-robot global transformation, intra-robot factor graph, and real-time pose estimation. First, a flexible load-aware bidirectional global pose optimization (FLB-GPO) module is designed to achieve inter-robot adaptive computational balance. Next, a multi-source evaluated local pose optimization (ME-LPO) module is introduced, which allows high-precision robots to assist low-precision robots for accurate intra-robot factor graph refinement on different scenes. Then, a geometric observation-driven degenerate pose correction (GO-DPC) module is developed to improve observation precision and real-time pose estimation accuracy in degraded environments. Together, these modules establish a hierarchical optimization architecture that achieves inter-robot computational balance, intra-robot refined pose, and real-time degeneration correction. Both simulation and real-world experiments demonstrate that the proposed framework consistently outperforms state-of-the-art methods in terms of computational balance and robustness.

Accurate estimation and compensation of extrinsic parameters (EPs) in vehicle-mounted integrated navigation systems are fundamental to achieving high-precision vehicle localization. In general, precise estimation of these parameters through information fusion methods, such as factor graph optimization (FGO), is only feasible when the vehicle motion provides sufficient excitation to render the parameters observable. However, during vehicle operation, many trajectories fail to excite certain EPs. Such weakly excited motion data not only reduces the estimation accuracy of EPs but also introduces unnecessary computational overhead in FGO. To address this issue, this study analyzes the observability of different EPs under various vehicle motion states, clarifying the excitation effects associated with specific motions. By fully incorporating all motion characteristics within the factor graph sliding window, a decoupled estimation framework is developed, which integrates observability constraints and EP convergence. Building on this framework, a lightweight and robust extrinsic estimation method is proposed to suppress the influence of weakly excited motions on unobservable parameters. Both simulations and field tests validate that, compared with other factor graph approaches, the proposed method effectively improves the estimation accuracy of EPs while reducing computational cost.

In urban scenarios such as logistics delivery and emergency rescue, unmanned aerial vehicles (UAVs) can obtain reliable positioning information through visual scene matching-based navigation to ensure successful task execution. However, traditional feature-based methods are susceptible to dynamic motion blur and significant scale variations in images, which often result in decreased positioning accuracy. In these critical applications, such failures can lead to catastrophic consequences ranging from mission failure and dangerous collisions to life-threatening delays in emergency rescue. To address these challenges, this paper proposes a visual positioning framework for UAVs based on salient position selection. In the front-end, an improved bilateral filtering algorithm, equipped with an adaptive parameter adjustment mechanism, is used to enhance image quality under challenging conditions, such as uneven illumination and motion blur. A salient position selection mechanism is introduced in the feature extraction network. It leverages channel attention, to dynamically enhance discriminative features, and applies spatial sparsity constraints, to focus on geometrically stable regions. As a result, the network extracts keypoints with distinctive structural features. After keypoint matching, the back-end constructs a unified confidence evaluation model that integrates feature matching quality and geometric consistency. Combined with a weighted sliding window optimization strategy, this enhances the positioning accuracy of UAV localization based on image matching. In real-world flight tests, the proposed method reduced localization error by 28.9%, compared to mainstream matching-based positioning algorithms. By leveraging only publicly available satellite imagery, it offers an effective solution for low-altitude UAV localization and navigation.

With the growing need for drones in complex and hazardous environments, this paper proposes a robust autonomous exploration method to address key limitations in existing algorithms, such as blind viewpoint selection, path redundancy, and excessive map nodes. Locally, a hybrid viewpoint generation strategy based on boundary geometry and space filling is introduced, along with a gain evaluation mechanism tailored to viewpoint types for improved path targeting. Globally, a dynamic viewpoint graph is constructed via branch extraction, with node pruning and edge constraints guided by historical and planning paths. A visibility-aware global path optimization further enhances smoothness and safety. Experiments show the proposed method reduces exploration time by approximately 12

Achieving reliable loop closing between aerial and ground robots is essential for heterogeneous multi-robot SLAM. However, existing methods often rely on homogeneous sensor configurations or are restricted to specific environments (either indoor or outdoor), limiting their generalizability across varying platforms and scenes. To address these challenges, this paper proposes a hierarchical density-enhanced binary and triangle combined descriptor with plane assistance for air-ground loop closing. First, an adaptive multi-frame fusion strategy is developed by incorporating sensor height, installation angle, and horizontal velocity, enabling rapid adaptation to diverse LiDARs on heterogeneous robots. Second, to address the indoor structural ambiguity, the plane-assisted ceiling-ground segmentation method is proposed. It can select ground while filtering out ceiling interference. Furthermore, to cope with air-ground viewpoint differences in outdoor scenes, we leverage a hierarchical density-enhanced BTC descriptor, and incorporate the incremental keypoints extraction and multi-layer density-guided loop detection for acceleration. Extensive benchmarks against state-of-the-art methods, such as GAPR, EHPR, and UniLGL, demonstrate that our method achieves superior performance on both indoor and outdoor air-ground datasets, while maintaining competitive generalization on cross-LiDAR multi-session datasets, with only a modest 2.9 ms runtime increase over BTC.

Underground sheltered spaces, such as utility tunnels, mines, and railways, are characterized by unique features including enclosure, complexity, and structural diversity, which present great challenges to collaborative perception and localization in clustered unmanned systems. This article first analyzes the current research on clustered unmanned systems in underground spaces. Second, considering the challenges of communication limitation, degraded environments, and constrained perception in underground spaces, this article elaborates on the progress of key technologies for collaborative localization and perception in unmanned systems, including data exchange methods for robotic cluster under communication constraints, robust and resilient simultaneous localization and mapping (SLAM) algorithms under degraded environments, collaboration strategies for heterogeneous multirobot systems in underground environments, and collaborative semantic SLAM under limited environmental perception. Finally, the technological bottlenecks and development trends of clustered unmanned systems in underground environments are summarized and discussed.

During unmanned aerial vehicle (UAV) operations, onboard imaging systems are highly susceptible to motion blur, which degrades critical image features and adversely affects the accuracy of subsequent target recognition and detection. To address this challenge, this paper proposes a novel neural network-based approach for direct recognition of long-range targets and scenes from motion-blurred images, eliminating the need for explicit image restoration. A datasets comprising images with varying degrees of motion blur was generated, and the YOLOv11 network was trained on this datasets to facilitate data augmentation and robust feature learning. Performance evaluation was conducted on images with various scenarios and different blur intensities. Experimental results demonstrate that the proposed approach significantly enhances target and scene recognition performance, even under severe motion blur conditions.

Space-time adaptive processing (STAP) serves as a potent tool for clutter suppression and moving target detection within airborne radar systems. Estimation of the clutter covariance matrix (CCM) stands as a pivotal challenge in STAP filter design and achieving sparse reconstruction of the clutter covariance matrix with a limited number of samples is of paramount importance. This paper presents a clutter suppression method tailored for airborne forward-looking array radar systems, grounded in principles of joint statistical analysis and structural prioritization. This approach facilitates the estimation of the clutter covariance matrix even in scenarios characterized by sample scarcity. Leveraging the assumption of adherence to an inverse Wishart prior distribution, the methodology derives a maximum posterior estimate by exploiting the intrinsic low-rank symmetry of the matrix. Simulation results, conducted using a radar forward-looking array model, demonstrate the efficacy of the proposed method in enhancing clutter suppression performance compared to traditional covariance matrix estimation techniques, all while maintaining computational efficiency.

The visual simultaneous location and mapping (VSLAM) method has been widely proposed to estimate the mobile equipment's position. However, in dim indoor application scenarios, there exists general image blur, which highly affects the precision and reliability of the VSLAM method. In this article, we proposed a novel pipeline named LTN-VINS based on a lightweight transformer network (LTN) and a self-adaptive point spread function (PSF) estimator, to improve VSLAM localization performance when motion blur occurs in images. First, based on modulated deformable convolution network (MDCN) and inertial measurement unit (IMU) pre-integration between image frames, a self-adaptive image PSF estimator (SAPE) is proposed to preprocess the input image. Second, considering that the feature point methods commonly used in most VSLAM systems are usually related to edge learning in images, we designed a differential high-pass filtering (DHF) module to further enhance the network's ability to extract image edges. Lastly, to lighten the computational burden and improve the network's ability to extract semantics between long-distance pixel dependencies, we propose a multihead linear self-attention (MLSA) mechanism. Extensive experiments in the public Euroc datasets and real-world underground parking environments demonstrate that the proposed method can effectively alleviate the image blur influence in VSLAM and achieve remarkable localization optimization performance compared to other SOTA deblurring network methods in terms of accuracy and robustness.

With the rapid development of the civil aviation industry, the reliability and real-time performance of airborne data transmission are becoming increasingly important. The traditional airborne network cannot meet the future flight requirements of the aircraft. To ensure the reliable and real-time transmission of data, the time-sensitive network introduces the Frame Replication and Elimination for Reliability (FRER) mechanism. The standard FRER mechanism defines the methods of frame replication and elimination of redundant frames. However, the description of how the replicated frames are transmitted is not in-depth. The frame replication and elimination function at the source and destination nodes will also reduce the reliability and real-time performance of the network. In order to realize the application of the time-sensitive network in the airborne network, this article independently builds an airborne time-sensitive network test simulation platform. It carries out in-depth research on improving the reliability of the network. It puts forward a path-finding algorithm based on a time-sensitive network with the FRER mechanism in response to the problem of low reliability of the selected data transmission paths in the airborne network. The algorithm integrates the constraints of transmission link delay and packet loss rate. It performs link reliability calculation before selecting redundant paths to obtain non-overlapping data transmission paths. The experimental results show that, compared with the dynamic link redundancy selection algorithm, the path delay is reduced by 21.51%. Compared with the multilevel P-cycle cascading algorithm, the path delay is reduced by 19.70%. At a 120 Mbps data transmission rate, the packet loss rate is reduced by 18.67% compared with the dynamic link redundancy selection algorithm. It is also reduced by 24.00% compared with the multilevel P-cycle cascading algorithm. These results show that the proposed method improves the reliability of data transmission in the airborne network.

To address the problems of insufficient utilization of complementary modal features and loss of detail information in infrared-visible image fusion, this paper proposes an infrared-visible image fusion method based on a Gated Depth-wise convolutional Transformer and Cascaded Cross-Attention. This method employs a Gated Depth-wise convolutional Transformer as the core module to construct an encoder-decoder framework. Efficient cross-modal fusion is achieved by enabling feature interaction across different levels during the fusion and decoding stages. At the encoder level, a dual-branch encoder processes the two modalities independently. Each encoder comprises a shallow convolutional feature extraction module and a multi-layer Transformer feature refinement module, extracting both shallow and deep features. At the fusion level, a cascaded gated cross attention guides the interaction between deep infrared and visible features within the Transformer space. Concurrently, skip connections with shallow features achieve alignment and fusion of cross-modal information between deep and shallow layers. At the decoder level, the same Gated Depth-wise convolutional Transformer module decodes the fused features. Shallow information is augmented via skip connections, and the feature maps are progressively decoded into the fused image through several convolutional layers for upsampling. To validate the performance of the algorithm, testing and verification were conducted using the TNO dataset. Test results demonstrate that the subjective visual quality and key objective metrics of the proposed method surpass those of mainstream methods such as CDDFuse, SwinFusion, and PSFusion. Furthermore, comprehensive ablation studies substantiate the optimality of the network architecture and its parameters.

In recent years, Wireless Sensor Networks (WSNs) have seen increasing application in GPS-Denied environments, such as indoor localization and mobile robot navigation. Among various positioning algorithms, the Least Squares Method is widely used for its simplicity and robustness. While numerous optimization algorithms have been developed to reduce localization error, comprehensive studies on the sources and propagation of positioning errors remain limited.This paper focuses on the error propagation mechanisms in range-based WSN localization. It conducts an in-depth analysis of the primary sources of positioning error, including ranging errors, anchor node position errors, and the influence of network topology on error propagation. A novel, simplified error propagation model based on covariance theory is proposed to quantitatively estimate the cumulative impact of these errors on localization accuracy.To evaluate the accuracy and practicality of the proposed algorithm, comprehensive assessments were carried out through both MATLAB simulations and physical experiments in representative indoor environments. The results show that the average discrepancy between the estimated and actual positioning errors remains within a few centimeters, demonstrating strong predictive capability. Furthermore, the Geometric Dilution of Precision is introduced to assess the influence of base station configurations. The experimental results further confirm that the proposed algorithm maintains strong robustness even in suboptimal geometric configurations of the network. In conclusion, the error propagation model proposed in this study not only significantly simplifies the theoretical computation of localization errors but also offers high accuracy in error estimation, making it particularly suitable for multi-sensor fusion scenarios that require precise error modeling. It also provides new insights and quantitative tools for optimizing WSN structures and improving overall localization performance.

This study introduces and validates a novel approach for estimating measurement errors and implementing data fusion technology. The proposed method addresses the challenge of diminished accuracy in fusion outcomes from inertial measurement units (IMU) array in harsh conditions or during IMUs malfunction. The second-order difference (SOMD) algorithm is improved in this method, and a novel adaptive weighting coefficient construction method is proposed by combining the measurement error level and its similarity. Experimental results indicate that, in comparison to the Adaptive Kalman Filtering method based on the SOMD algorithm and the Weighted Average method, the method proposed in this paper enhances the suppression of multi-source noise such as zero-bias stability, zero-bias instability, and angle random walk in gyroscopes and accelerometers by over 40

This study provides both qualitative and quantitative comparisons of wave and ray optics simulations for long-exposure incoherent imaging in weak to strong atmospheric turbulence. The ray tracing image shows a recognizable approximation to the wave optics results even in the saturation regime, but it is advantageous in fast computational speed and sampling unconstraint. The analytic expression of the ray-based Fried parameter is also proposed for the first time to predict the spatial-resolution effects of turbulence on imaging and beam propagation of geometric optics, offering quantitative guidance for the application of ray optics to approximate the wave optics results.