Search Results (713)

Sixth-generation (6G) vehicular systems are expected to provide extreme reliability, ultra-low latency, and high data rates while simultaneously enabling accurate and timely perception of vehicles, road users, and surrounding environments. Conventional vehicle-to-everything (V2X) systems treat communication and environmental sensing separately and therefore fall short in dynamic, dense, and safety-critical driving scenarios. Integrated sensing and communication for V2X (ISAC-V2X) addresses this limitation by sharing hardware and spectrum for dual-functional operation. This survey consolidates the state of the art across eight thematic areas: fundamentals and taxonomy; vehicular channel models and sensing metrics; physical-layer techniques, including waveform design, beamforming, reconfigurable intelligent surfaces, non-orthogonal multiple access, and rate-splitting multiple access; security and privacy; networked and cell-free ISAC with multi-access edge computing; multi-modal perception using radar, LiDAR, camera, and radio-frequency data; artificial intelligence and ML for channel, beam, target, and sensing-data processing; and future research directions.Fiveconsolidated research gaps are synthesized from the reviewed literature: end-to-end frameworks for networked ISAC-V2X under high mobility; unified multi-modal AI-native architectures for joint sensing, communication, localization, and security; incomplete standardization and interoperability; limited validation, testbeds, and reproducible benchmarks; and scalability, robustness, and cross-domain optimization. This survey concludes with standardization and testbed recommendations and presents a roadmap toward deployable 6G ISAC-V2X systems. Full article

High-precision manufacturing in unstructured factories imposes stringent requirements on real-time scene perception and end-effector positioning accuracy. Traditional single-sensor solutions suffer from perception blind spots in human-robot mixed environments with complex lighting, while chassis cumulative error often leads to rigid collisions during end-effector operations. To address this, this paper proposes and evaluates a decision-level multi-sensor coordination mechanism for robust navigation and high-precision planar positioning of industrial mobile robots. The mechanism assigns explicit sensor roles, distance-dependent trigger conditions, and deterministic safety priorities. At the navigation and obstacle avoidance level, a sequential decision policy is constructed: macroscopically, a lightweight You Only Look Once version 5 small (YOLOv5s) is utilized for the early detection of dynamic objects, providing bounding-box coordinates to trigger preemptive deceleration, while LiDAR independently provides geometric ranging for ROS local-costmap updating and detour replanning; microscopically, a low-level hardware interrupt strategy triggered by ultrasonic sensors is proposed to mitigate near-field blind spots and reduce communication latency. At the end-effector positioning level, under illumination conditions ranging from 200 to 1000 lux, an adaptive alignment algorithm combining hue-saturation-value color-space morphological processing and Kalman filtering is proposed to suppress measurement noise caused by illumination variations and mechanical vibrations. Experiments in the tested dynamic human-robot mixed scenarios showed no rigid collisions for the proposed system and an emergency response time of approximately 50 ms against sudden blind-spot intrusions. Simultaneously, the system achieves a 95% reliability rate in controlling the end-effector 2D planar positioning error (X-Y plane) within a ±2 mm tolerance under complex illumination interference. These results demonstrate improved navigation safety and planar-positioning reliability under the tested flexible-manufacturing conditions. Full article

Bionic intelligent robots operating in unstructured dynamic environments require perception systems that regulate uncertain observations according to their reliability and avoid converting transient disturbances into persistent spatial references. Inspired by reliability-weighted multisensory integration and by a functional abstraction of persistence-based evidence consolidation in biological navigation, this paper proposes a bio-inspired reliability-constrained LiDAR–inertial odometry framework. Each point-to-map observation is evaluated using residual consistency, local geometric quality, and voxel-level temporal stability. The fused reliability score regulates both the ESIKF state update and incremental map maintenance: low-confidence observations are continuously down-weighted, highly reliable points are admitted to the persistent map, ambiguous points are retained in short-term candidate memory for multi-frame verification, and unreliable points are rejected. The framework translates biological design principles into an engineering perception–memory architecture rather than reproducing a specific neural circuit. Repeated experiments on public datasets and a wheeled mobile robot platform show comparable accuracy in two normal sequences. Across five dynamic sequences, the complete method reduces localization RMSE by 14.08–28.88% relative to Fast-LIO2 and achieves lower mean RMSE than Dynamic-LIO on all five evaluated dynamic sequences. Frozen-map evaluation further yields 7.77–15.71% lower point-to-map RMSE together with higher consistent-correspondence ratios and coverage, while the maximum mean RMSE deviation in the parameter-sensitivity study remains below 7.2%. The maximum measured processing time is 18.69 ms per scan, maintaining real-time operation for a 10 Hz LiDAR. Full article

Dense partial-view LiDAR observations are attractive for outdoor perception, but limited overlap and viewpoint sensitivity make odometry and mapping less reliable than with spinning LiDARs. Many recent algorithms for this sensing regime are built as LiDAR-inertial odometry frameworks, whose localization and mapping performance can degrade or fail when the IMU state estimation becomes unstable. This paper presents a LiDAR-only framework for complex outdoor scenes using a factor-graph back-end. After denoising and motion compensation, the point cloud is projected onto a range image for ground, planar, edge, and line extraction. Pose estimation is strengthened by degeneracy-aware feature selection, while loop closing combines scan-based and path-based cues to handle partial-view revisits. Experiments in tunnels, urban roads, residential areas, and other challenging scenes show reduced drift and improved mapping consistency for dense limited-FoV LiDAR data. Full article

To address the challenges of insufficient feature representation and the difficulty of detecting sparse and distant objects in UAV-borne LiDAR point clouds—which exhibit significantly lower point density than terrestrial/mobile LiDAR scans—this paper proposes an enhanced detection algorithm built upon the PointPillars framework. First, a coordinate attention mechanism is incorporated to enhance the network’s ability to capture spatial geometric information. Furthermore, the backbone network is redesigned with a dual-path structure and a feature modulation fusion module, enabling adaptive integration of multi-scale features. Experimental evaluations conducted on a custom simulated UAV-borne LiDAR point cloud dataset demonstrate that the proposed method achieves 85.35% 3D mAP and 88.26% BEV mAP, corresponding to absolute improvements of 21.61 and 7.85 percentage points compared with the original PointPillars model. In addition, the proposed approach demonstrates consistent performance on the publicly available KITTI benchmark through preliminary cross-dataset validation. The results indicate that the proposed method can effectively improve the detection accuracy and robustness of LiDAR-based 3D object detection in complex environments. Full article

This work presents the design, construction and operation of the SKD-1, a modular skid-steer unmanned ground vehicle (UGV) developed as a low-cost research platform for mobile robotics applications. The platform integrates a differential skid-steer drive system, a Raspberry Pi-based onboard computer, and a microcontroller-based control layer implemented using an STM32 microcontroller. The sensing system includes light detection and ranging (LiDAR), global navigation satellite system (GNSS), and an inertial measurement unit (IMU), enabling experiments in localization, mapping, and autonomous navigation. The software architecture is based on the Robot Operating System (ROS) 2 framework, relying on standard ROS 2 packages for perception, mapping, and path planning, with the custom hardware-interface layer being the only non-standard software component. The mechanical and electronic subsystems were designed with a modular architecture that facilitates maintenance, sensor replacement, and hardware upgrades. The primary contribution of this work is the open-hardware design, integration, and documentation of a reproducible robotics testbed, motivated by the prohibitive cost of commercial platforms in resource-constrained research contexts. Indoor and outdoor experiments—covering velocity-tracking, SLAM, and waypoint-navigation trials—demonstrate the functional integration of the sensing, actuation, and computing subsystems. Full article

Road markings support lane guidance, traffic regulation, and machine perception, but their field inspection still depends largely on manual surveys or local retroreflectivity measurements. This study presents a vehicle-mounted inspection framework that fuses LiDAR intensity with camera-based semantic segmentation for automated road marking wear assessment. The platform integrates LiDAR, a stereo camera, IMU, GNSS, and an industrial computer. LiDAR-inertial mapping provides spatial alignment, while ground filtering, region-of-interest extraction, and adaptive intensity thresholding generate preliminary marking candidates. A YOLOv11 segmentation model produces pixel-level marking masks, and LiDAR candidates are projected onto the image plane for semantic confirmation. Confirmed points are accumulated into grid cells and evaluated using reflectance, point density, fusion retention, and geometric coverage indicators. In a representative route, 2441 grid units were analyzed: 1210 units were valid for formal grading, with 1060 good, 131 slightly worn, and 19 moderately worn units; 1231 units were reserved for review. The mean composite score of the valid grids was 0.906. A five-report aggregate further showed that 87.2% of the units received valid wear categories. The results indicate that multimodal fusion transforms road marking inspection into a quantitative, spatially referenced, and reportable process. Full article

Accurate localization and mapping are essential for autonomous mobile robots operating in unknown environments. This study investigates the impact of Extended Kalman Filter (EKF)-based sensor fusion on the performance of three widely used two-dimensional (2D) LiDAR Simultaneous Localization and Mapping (SLAM) algorithms: GMapping, Karto SLAM, and SLAM Toolbox. Wheel encoder longitudinal velocity and inertial measurement unit (IMU) yaw angular velocity were fused using an EKF and compared with raw wheel odometry using the MIT Stata Center dataset. Localization performance was evaluated both before and after SLAM using translational and rotational Absolute Pose Error (APE) across multiple trajectory segments. Five repeated executions were performed for each SLAM configuration to characterize run-to-run variability. Prior to SLAM, EKF-filtered odometry reduced translational APE root mean square error (RMSE) by approximately 61–75% and rotational APE RMSE by approximately 65–77% relative to raw odometry. After SLAM, translational differences between the two odometry sources were substantially smaller and varied according to the evaluated algorithm and trajectory, while rotational performance exhibited larger and less consistent changes. These results demonstrate that substantial improvements in upstream odometry accuracy do not necessarily produce proportional improvements in final SLAM localization and that the influence of sensor fusion varied across the evaluated SLAM algorithm and trajectory segments, providing practical guidance for selecting localization strategies in autonomous mobile robots. Full article

Automated three-dimensional (3D) mapping with a survey-grade terrestrial laser scanner (TLS) is the canonical instance of robotic stop-and-scan light detection and ranging (LiDAR) mapping: the mobile platform must remain stationary for minutes per scan, so the environment must be covered with as few scans as possible. Where to stop hinges on a coverage model predicting what a scan pose will observe. The conventional isotropic-disk model counts every free cell within sensing range as covered—including cells behind walls—and, therefore, skips scans that genuine observation requires. We replace the disk with a ray-cast visibility region computed on the robot’s live two-dimensional (2D) occupancy map, admitting only cells in direct line of sight; define a line-of-sight (LOS) coverage metric over mapped free space and drive a marginal-gain scan/skip rule embedded in frontier exploration. The planner thus performs 2D scan-station placement for subsequent 3D mapping. In a controlled paired ablation in simulation, the visibility rule raises LOS coverage from about 77% to 84–85% for one to two additional scans, and it also outperforms a disk baseline governed by the identical marginal-gain rule, isolating the coverage model itself as the decisive factor. In a fully autonomous on-hardware comparison in an industrial test room, with each rule driving a mobile platform carrying a Leica BLK360 G1, the visibility rule raised LOS coverage from 77.6% to 92.1%; every visibility run’s scans registered offline at survey grade (4–5 mm bundle error, 84–88% overlap), whereas disk runs yielded at best a single-link network and once a single unregistrable scan. The results indicate that, for the room-scale indoor environments studied, occlusion-aware visibility is a sounder basis than Euclidean proximity for stop-and-scan placement. Full article

UAVs increasingly rely on accurate SLAM for aerial mapping and inspection in GPS-denied environments. 3D Gaussian Splatting (3DGS) has opened a new direction for UAV mapping by allowing SLAM systems to build dense, photorealistic, and renderable maps. Yet in 3DGS SLAM the map is optimized from the pose graph, so a false loop closure can deform both the UAV trajectory and the Gaussian map consumed by downstream UAV autonomy. Reliable loop admission is therefore relevant to safe GPS-denied operation because it protects the state and map estimates on which autonomous functions depend. The present work evaluated this upstream estimation-integrity problem; it did not measure closed-loop guidance, control, or navigation-safety outcomes. We address the loop-admission problem that arises after a place-recognition (PR) module proposes a candidate loop and relative-pose seed. GSSeq is a rendered-reference sequential verifier that uses the current Gaussian map as active evidence before inserting a loop factor. It renders RGB-D references with the PR seed, checks LiDAR/rendered-depth consistency and image/rendered-reference consistency over active support, and propagates the seed through a short query trajectory window. A loop is admitted only when this evidence remains geometrically supported and photometrically stable. On fixed LiDAR-PR candidate sets spanning MARS-LVIG, MUN-FRL, and independent NTU-VIRAL aerial sequences together with ground-mobility benchmarks, GSSeq provides a competitive precision-oriented operating point while suppressing false loop admissions. Thresholds calibrated only on NTU-VIRAL spms_01 combine rendered RGB agreement with LiDAR-submap geometry and are then frozen for spms_02. On this held-out sequence, GSSeq rejects all seven false-positive BTC factors while retaining one of three true-positive factors. The trajectory-to-map experiment reduced ATE RMSE from $2.609\mathrm{m}$ to $1.417\mathrm{m}$ and improved selected-view PSNR from $13.80\mathrm{dB}$ to $16.46\mathrm{dB}$. These results show that rendered verification can preserve an aligned, renderable UAV trajectory-map pair before unsupported loop factors reshape the SLAM map. Full article

The three-dimensional documentation of crash-damaged vehicles can support the visual and geometric recording of deformation, but it is unclear how complete mobile reconstruction workflows compare when applied to the same vehicles. This study compared three workflows using a single consumer device, an Apple iPhone 16 Pro Max: reconstruction from photographs, reconstruction from extracted video frames, and direct mobile light detection and ranging (LiDAR) scanning. Three damaged vehicles were documented: a Volkswagen Passat B6 Variant, a Toyota Auris, and a Toyota Yaris. RealityScan was used for reconstruction from photographs and video frames, and Polycam was used for the LiDAR scans. In CloudCompare, all models were cleaned, scaled using the known wheelbase, registered to the LiDAR reference by the Iterative Closest Point algorithm, and compared using cloud-to-mesh distances, with the principal quantitative statistics based on absolute point-to-surface distance magnitudes. Because the mobile LiDAR model served as an internal reference rather than as an independent metrological ground truth, the reported values describe residual post-registration point-to-surface deviations and not absolute geometric accuracy. The principal surface evaluation used exactly 100,000 surface-sampled points per evaluated direction and bidirectional cloud-to-mesh calculations. The standardised results did not show a uniform ordering between reconstruction from photographs and reconstruction from video frames. In the reconstruction-to-LiDAR direction, median absolute distances ranged from 0.03082 to 0.03677 m for the Passat, from 0.02900 to 0.03413 m for the Auris, and from 0.05953 to 0.06168 m for the Yaris. Lower reverse-direction median values and the broader upper-tail distributions observed for the Yaris demonstrated the directional character of the surface comparison. The Yaris showed larger, long-tailed deviations concentrated mainly in the rear and left-lateral damaged regions. However, because each damage configuration was represented by only one vehicle, the observed differences cannot be attributed to damage type alone. The three workflows provided complementary geometric and visual information for crash-damaged vehicle documentation, although model fusion and accident-reconstruction parameters were not evaluated in this study. Because only one acquisition was performed for each vehicle–workflow combination, the findings should be interpreted as an exploratory comparison rather than as an assessment of repeatability, operator variability, or measurement uncertainty. Full article

Reliable aboveground biomass (AGB) estimates for woody crops are essential for carbon accounting and for Measurement, Reporting and Verification (MRV) frameworks. However, it remains unclear how LiDAR modality and sampling geometry influence plot-scale and tree-scale AGB predictions in intensively managed Mediterranean orchards. In this study, we benchmarked four LiDAR modalities, namely open national airborne laser scanning from the Spanish National Aerial Orthophotography Plan (PNOA/ALS), a dedicated Riegl airborne laser scanner (ALS), unmanned laser scanning (ULS) and mobile laser scanning (MLS), across three woody-crop sites in Córdoba (southern Spain): IFAPA, Doña María, and Villaseca. Plot-level LiDAR metrics (mean height, 95th height percentile, maximum height, and canopy-cover proxies) were extracted from normalized point clouds and related to field AGB using Random Forest and XGBoost regression models, together with an ensemble predictor, under an 80/20 train–test split. In parallel, TreeQSM-based Quantitative Structure Models (QSMs) were evaluated as an independent tree-level three-dimensional reconstruction approach. XGBoost achieved the lowest errors at IFAPA (RMSE = 0.400 Mg ha⁻¹; R² = 0.994) and Villaseca (RMSE = 0.872 Mg ha⁻¹; R² = 0.995), whereas PNOA/ALS was competitive at Doña María (RMSE = 0.725 Mg ha⁻¹; R² = 0.994). TreeQSM closely matched the field inventory at the low-biomass IFAPA site but tended to overestimate biomass at Doña María and Villaseca, and only 28% of scanned trees yielded usable reconstructions. The results support the use of cross-platform LiDAR for orchard AGB and carbon mapping and identify the conditions under which open national LiDAR can enable scalable MRV of Mediterranean woody crops. Full article

Integrated qiaopi office–residence heritage preserves the material setting of remittance-letter operations together with domestic, educational, and ritual activities. In Lingnan’s hot–humid and disaster-prone environment, condition records need to be repeatable, spatially traceable, and continuously updatable. Taking Jingzu Jiashu and Mingde Jiashu, two former qiaopi office sites in Chaoshan, as case studies, this research develops an evidence-traceable digital conservation workflow integrating multi-source documentation; an adopted YOLOv8 surface-deterioration baseline; qualitative Grad-CAM visualization; structured deterioration records; and semi-automatic, human-confirmed Revit/HBIM association. UAV and terrestrial photography, mobile LiDAR/scanning, handheld measurement, measured drawings, point-cloud and reality-based products, and geometric models were organized into case-specific HBIM environments. The adopted deterioration dataset comprised 362 original images at 512 × 512 pixels and 2024 bounding-box annotations for five visually identifiable categories: spalling, staining, plants, saltpetering, and crack. The original images were divided into 253 training, 72 validation, and 37 independent-test images, while augmentation was restricted to the training subset, increasing the training pool to 1600 images. The previously established YOLOv8 baseline achieved a Precision of 0.85, Recall of 0.72, mAP50 of 0.83, and mAP50–95 of 0.58. Grad-CAM heatmaps were used as qualitative aids to examine model-emphasized image regions. Retained detections associated with Jingzu Jiashu and Mingde Jiashu were converted into versioned records containing source-image identifiers, deterioration classes, detector confidence, survey information, spatial references, verification states, and revision histories. Candidate spatial associations were generated through case identifiers, façade or space zones, element identifiers, and available spatial evidence, while final M1–M3 associations required human confirmation. By preserving source provenance, spatial uncertainty, and record histories, the workflow provides an auditable information basis for routine inspection, post-event review, maintenance prioritization, repair interpretation, and resilience-oriented preventive conservation. The workflow supports screening-level deterioration recognition and information management but does not provide causal diagnosis, structural assessment, exact affected-area measurement, building-independent generalization, or automatic repair recommendations. Full article

Existing reviews on AI in tea production are either agriculture-generic or limited to isolated tasks. This review thoroughly compares vision technologies (RGB, hyperspectral, near-infrared, thermal, Light Detection and Ranging (LiDAR), Unmanned Aerial Vehicle (UAV)) and establishes a task-oriented algorithm selection framework for the tea industry. For small-sample or near-linear problems, traditional machine learning (ML) (support vector machine (SVM); partial least squares regression (PLSR)) remains effective. For unstructured field tasks, deep learning achieves superior performance: pest detection accuracy exceeds 97%, tea bud detection reaches 96.8% with RGB images, and hyperspectral imaging predicts nitrogen content with R² > 0.90 and tea polyphenols with R² up to 0.925. Algorithm choice further differentiates by task granularity: lightweight convolutional neural networks (CNNs) balance speed and accuracy for edge deployment at 16 fps; You Only Look Once (YOLO) series detectors enable real-time localization on mobile platforms at 93.1% accuracy, 24 ms per target. No single algorithm dominates all tea tasks; selection is a trade-off among accuracy, speed, data availability, and computational constraints. These findings outline a structured analysis of the challenges and pathways for transitioning computer vision (CV) from laboratory research toward field-deployable tools. Full article

Mobile LiDAR systems (MLS) enable rapid acquisition of large-scale 3D point cloud data. Semantic segmentation of the acquired point clouds is an important task in outdoor scene understanding and environmental perception for autonomous driving. However, existing methods tend to suffer from boundary confusion when segmenting MLS point clouds with long-tail categories. To address this problem, we propose the Dilated Context Attention Network (DCA-Net), which consists of a dilated local geometric encoding module, a channel attention pooling module, and a category-boundary sampling strategy. The dilated local geometric encoding module expands point-to-point connections within a fixed neighborhood to strengthen contextual modeling among neighboring points. The channel attention pooling module uses a channel attention mechanism to enhance informative channel responses in neighborhood features, thereby improving local feature representation. The category-boundary sampling strategy increases the sampling probabilities of minority-category points and boundary points, reducing feature information loss during down-sampling. Experimental results on the S3DIS, Toronto3D, and MLS road scene datasets show that DCA-Net achieves mIoU scores of 69.4%, 84.1%, and 96.6%, respectively. These results demonstrate that the proposed method alleviates boundary confusion in point cloud segmentation with long-tail categories, without causing a noticeable degradation in the segmentation performance of majority categories. Full article