Search Results (1,186)

Environmental perception is crucial for safe autonomous driving, enabling accurate analysis of the vehicle’s surroundings. While 3D LiDAR is traditionally used for 3D environment reconstruction, its high cost and complexity present challenges. In contrast, camera-based cross-view frameworks can offer a cost-effective alternative. Hence, this manuscript proposes a new cross-view model to extract mapping features from camera images and then transfer them to a Bird’s-Eye View (BEV) map. Particularly, a multi-head attention mechanism in the decoder architecture generates the final semantic map. Each camera learns embedding information corresponding to its position and angle within the BEV map. Cross-view attention fuses information from different perspectives to predict top-down map features enriched with spatial information. The multi-head attention mechanism then globally performs dependency matches, enhancing long-range information and capturing latent relationships between features. Transposed convolution replaces traditional upsampling methods, avoiding high similarities of local features and facilitating semantic segmentation inference of the BEV map. Finally, we conduct numerous simulation experiments to verify the performance of our cross-view model. Full article

Point cloud registration is pivotal across various applications, yet traditional methods rely on unordered point clouds, leading to significant challenges in terms of computational complexity and feature richness. These methods often use k-nearest neighbors (KNN) or neighborhood ball queries to access local neighborhood information, which is not only computationally intensive but also confines the analysis within the object’s boundary, making it difficult to determine if points are precisely on the boundary using local features alone. This indicates a lack of sufficient local feature richness. In this paper, we propose a novel registration strategy utilizing ordered point clouds, which are now obtainable through advanced depth cameras, 3D sensors, and structured light-based 3D reconstruction. Our approach eliminates the need for computationally expensive KNN queries by leveraging the inherent ordering of points, significantly reducing processing time; extracts local features by utilizing 2D coordinates, providing richer features compared to traditional methods, which are constrained by object boundaries; compares feature similarity between two point clouds without keypoint extraction, enhancing efficiency and accuracy; and integrates image feature-matching techniques, leveraging the coordinate correspondence between 2D images and 3D-ordered point clouds. Experiments on both synthetic and real-world datasets, including indoor and industrial environments, demonstrate that our algorithm achieves an optimal balance between registration accuracy and efficiency, with registration times consistently under one second. Full article

Objective: To develop a non-invasive cognitive impairment detection system using speech data analysis, addressing the growing global dementia crisis and enabling accessible early screening through daily health monitoring. Methods: Speech data from 223 Korean patients were collected across eight tasks. Patients were classified based on Korean Mini-Mental State Examination scores. Four machine learning models were tested for three binary classification tasks. Voice acoustic features were extracted and analyzed. Results: The Deep Neural Network model performed best in two classification tasks, with Precision-Recall Area Under the Curve scores of 0.737 for severe vs. no impairment and 0.726 for mild vs. no impairment, while Random Forest achieved 0.715 for severe + mild vs. no impairment. Several acoustic features emerged as potentially important indicators, with DDA shimmer from the /i/ task and stdevF0 from the /puh-tuh-kuh/ task showing consistent patterns across classification tasks. Conclusions: This preliminary study suggests that certain acoustic features may be associated with cognitive status, though demographic factors significantly influence these relationships. Further research with demographically matched populations is needed to validate these findings. Full article

The automatic extraction of large-scale tie points (TPs) for Synthetic Aperture Radar (SAR) images is crucial for generating SAR Digital Orthophoto Maps (DOMs). This task involves matching SAR images under various conditions, such as different resolutions, incidence angles, and orbital directions, which is highly challenging. To address the feature extraction challenges of different SAR images, we propose a Gamma Modulated Phase Congruency (GMPC) model. This improved phase congruency model is defined by a Gamma Modulation Filter (GMF) and an adaptive noise model. Additionally, to reduce layover interference in SAR images, we introduce a GMPC-Harris feature point extraction method with layover perception. We also propose a matching method based on the SAR Modality Independent Neighborhood Fusion (SAR-MINF) descriptor, which fuses feature information from different neighborhoods. Furthermore, we present a graph-based overlap extraction algorithm and establish an automated workflow for large-scale TP extraction. Experiments show that the proposed SAR-MINF matching method increases the Correct Match Rate (CMR) by an average of 31.2% and the matching accuracy by an average of 57.8% compared with other prevalent SAR image matching algorithms. The proposed TP extraction algorithm can extract full-degree TPs with an accuracy of less than 0.5 pixels for more than 98% of 2-degree TPs and over 95% of multidegree TPs, meeting the requirements of DOM production. Full article

In tasks such as wood defect repair and the production of high-end wooden furniture, ensuring the consistency of the texture in repaired or jointed areas is crucial. This paper proposes the WTSM-SiameseNet model for wood-texture-similarity matching and introduces several improvements to address the issues present in traditional methods. First, to address the issue that fixed receptive fields cannot adapt to textures of different sizes, a multi-receptive field fusion feature extraction network was designed. This allows the model to autonomously select the optimal receptive field, enhancing its flexibility and accuracy when handling wood textures at different scales. Secondly, the interdependencies between layers in traditional serial attention mechanisms limit performance. To address this, a concurrent attention mechanism was designed, which reduces interlayer interference by using a dual-stream parallel structure that enhances the ability to capture features. Furthermore, to overcome the issues of existing feature fusion methods that disrupt spatial structure and lack interpretability, this study proposes a feature fusion method based on feature correlation. This approach not only preserves the spatial structure of texture features but also improves the interpretability and stability of the fused features and the model. Finally, by introducing depthwise separable convolutions, the issue of a large number of model parameters is addressed, significantly improving training efficiency while maintaining model performance. Experiments were conducted using a wood texture similarity dataset consisting of 7588 image pairs. The results show that WTSM-SiameseNet achieved an accuracy of 96.67% on the test set, representing a 12.91% improvement in accuracy and a 14.21% improvement in precision compared to the pre-improved SiameseNet. Compared to CS-SiameseNet, accuracy increased by 2.86%, and precision improved by 6.58%. Full article

Airborne LiDAR (ALS) and terrestrial LiDAR (TLS) data integration provides complementary perspectives for acquiring detailed 3D forest information. However, challenges in registration arise due to feature instability, low overlap, and differences in cross-platform point cloud density. To address these issues, this study proposes an automatic point cloud registration method based on the consistency of the single-tree position distribution in multi-species and complex forest scenes. In this method, single-tree positions are extracted as feature points using the Stepwise Multi-Form Fitting (SMF) technique. A novel feature point matching method is proposed by constructing a polar coordinate system, which achieves fast horizontal registration. Then, the Z-axis translation is determined through the integration of Cloth Simulation Filtering (CSF) and grid-based methods. Finally, the Iterative Closest Point (ICP) algorithm is employed to perform fine registration. The experimental results demonstrate that the method achieves high registration accuracy across four forest plots of varying complexity, with root-mean-square errors of 0.0423 m, 0.0348 m, 0.0313 m, and 0.0531 m. The registration accuracy is significantly improved compared to existing methods, and the time efficiency is enhanced by an average of 90%. This method offers robust and accurate registration performance in complex and diverse forest environments. Full article

To address the issue of significant point cloud ghosting in the construction of high-precision point cloud maps by low-speed intelligent mobile vehicles due to the presence of numerous dynamic obstacles in the environment, which affects the accuracy of map construction, this paper proposes a LiDAR-based Simultaneous Localization and Mapping (SLAM) algorithm tailored for dynamic scenes. The algorithm employs a tightly coupled SLAM framework integrating LiDAR and inertial measurement unit (IMU). In the process of dynamic obstacle removal, the point cloud data is first gridded. To more comprehensively represent the point cloud information, the point cloud within the perception area is linearly discretized by height to obtain the distribution of the point cloud at different height layers, which is then encoded to construct a linear discretized height descriptor for dynamic region extraction. To preserve more static feature points without altering the original point cloud, the Random Sample Consensus (RANSAC) ground fitting algorithm is employed to fit and segment the ground point cloud within the dynamic regions, followed by the removal of dynamic obstacles. Finally, accurate point cloud poses are obtained through static feature matching. The proposed algorithm has been validated using open-source datasets and self-collected campus datasets. The results demonstrate that the algorithm improves dynamic point cloud removal accuracy by 12.3% compared to the ERASOR algorithm and enhances overall mapping and localization accuracy by 8.3% compared to the LIO-SAM algorithm, thereby providing a reliable environmental description for intelligent mobile vehicles. Full article

The growing demand for high-accuracy mapping and 3D modeling using unmanned aerial vehicles (UAVs) has accelerated advancements in flight dynamics, positioning accuracy, and imaging technology. Structure from motion (SfM), a computer vision-based approach, is increasingly replacing traditional photogrammetry through facilitating the automation of processes such as aerial triangulation (AT), terrain modeling, and orthomosaic generation. This study examines methods to enhance the accuracy of SfM-based AT through real-time kinematic (RTK) UAV imagery, focusing on large-scale infrastructure applications, including a dam and its entire basin. The target area, primarily consisting of homogeneous water surfaces, poses considerable challenges for feature point extraction and image matching, which are crucial for effective SfM. To overcome these challenges and improve the AT accuracy, a constraint equation was applied, incorporating weighted 3D coordinates derived from RTK UAV data. Furthermore, oblique images were combined with nadir images to stabilize AT, and confidence-based filtering was applied to point clouds to enhance geometric quality. The results indicate that assigning appropriate weights to 3D coordinates and incorporating oblique imagery significantly improve the AT accuracy. This approach presents promising advancements for RTK UAV-based AT in SfM-challenging, large-scale environments, thus supporting more efficient and precise mapping applications. Full article

The automatic registration of infrared and visible images in vehicular imaging systems remains challenging in vision-assisted driving systems because of differences in imaging mechanisms. Existing registration methods often fail to accurately register infrared and visible images in vehicular imaging systems due to numerous spurious points during feature extraction, unstable feature descriptions, and low feature matching efficiency. To address these issues, a robust and efficient registration of infrared and visible images for vehicular imaging systems is proposed. In the feature extraction stage, we propose a structural similarity point extractor (SSPE) that extracts feature points using the structural similarity between weighted phase congruency (PC) maps and gradient magnitude (GM) maps. This approach effectively suppresses invalid feature points while ensuring the extraction of stable and reliable ones. In the feature description stage, we design a rotation-invariant feature descriptor (RIFD) that comprehensively describes the attributes of feature points, thereby enhancing their discriminative power. In the feature matching stage, we propose an effective coarse-to-fine matching strategy (EC2F) that improves the matching efficiency through nearest neighbor matching and threshold-based fast sample consensus (FSC), while improving registration accuracy through coordinate-based iterative optimization. Registration experiments on public datasets and a self-established dataset demonstrate the superior performance of our proposed method, and also confirm its effectiveness in real vehicular environments. Full article

In recent years, the rise of unmanned technology has made Simultaneous Localization and Mapping (SLAM) algorithms a focal point of research in the field of robotics. SLAM algorithms are primarily categorized into visual SLAM and laser SLAM, based on the type of external sensors employed. Laser SLAM algorithms have become essential in robotics and autonomous driving due to their insensitivity to lighting conditions, precise distance measurements, and ease of generating navigation maps. Throughout the development of SLAM technology, numerous effective algorithms have been introduced. However, existing algorithms still encounter challenges, such as localization errors and suboptimal utilization of sensor data. To address these issues, this paper proposes a tightly coupled SLAM algorithm based on similarity detection. The algorithm integrates Inertial Measurement Unit (IMU) and LiDAR odometry modules, employs a tightly coupled processing approach for sensor data, and utilizes curvature feature optimization extraction methods to enhance the accuracy and robustness of inter-frame matching. Additionally, the algorithm incorporates a local keyframe sliding window method and introduces a similarity detection mechanism, which reduces the real-time computational load and improves efficiency. Experimental results demonstrate that the algorithm achieves superior performance, with reduced positioning errors and enhanced global consistency, in tests conducted on the KITTI dataset. The accuracy of the real trajectory data compared to the ground truth is evaluated using metrics such as ATE (absolute trajectory error) and RMSE (root mean square error). Full article

Metabolomic data often present challenges due to high dimensionality, collinearity, and variability in metabolite concentrations. Machine learning (ML) application in metabolomic analyses is enabling the extraction of meaningful information from complex data. Bringing together domain-specific knowledge from metabolomics with explainable ML methods can refine the predictive performance and interpretability of models used in atherosclerosis research. In this work, we aimed to identify the most impactful metabolites associated with the presence of atherosclerotic cardiovascular disease (ASCVD) in cross-sectional case–control studies using explainable ML methods integrated with metabolomics domain knowledge. For this, a subset from the FLEMENGHO cohort with metabolomic data available was used as the training cohort, including 63 patients with a history of ASCVD and 52 non-smoking controls matched by age, sex, and body mass index from the same population. First, Partial Least Squares Discriminant Analysis (PLS-DA) was applied for dimensionality reduction. The selected metabolites’ correlations were analyzed by considering their chemical categorization. Then, eXtreme Gradient Boosting (XGBoost) was used to identify metabolites that characterize ASCVD. Next, the selected metabolites were evaluated in an external cohort to determine their effectiveness in distinguishing between cases and controls. A total of 56 metabolites were selected for ASCVD discrimination using PLS-DA. The primary identified metabolites’ superclasses included lipids, organic acids, and organic oxygen compounds. Upon integrating these metabolites with the XGBoost model, the classification yielded a test area under the curve (AUC) of 0.75. SHAP analyses ranked cholesterol, 3-methylhistidine, and glucuronic acid among the most impactful features and showed the diversity of metabolites considered for building the ASCVD discriminator. Also using XGBoost, the selected metabolites achieved an AUC of 0.93 in an independent external validation cohort. In conclusion, the combination of different metabolites has the potential to build classifiers for ASCVD. Integrating metabolite categorization within the SHAP analysis further enhanced the interpretability of the model, offering insights into metabolite-specific contributions to ASCVD risk. Full article

The ICESat-2 Land and Vegetation Height (ATL08) product is a new control point dataset for large-scale topographic mapping and geodetic surveying. However, its elevation accuracy is typically affected by multiple factors. The study aims to propose a new approach to classify ATL08 terrain estimates into different accuracy levels and extract reliable ground control points (GCPs) from ICESat-2 ATL08. Specifically, the methodology is divided into three stages. First, the ATL08 terrain estimates are matched with the raw ATL03 photon cloud data, and the ATL08 terrain estimates are used to fit a continuous terrain curve. Then, using the fitted continuous terrain curve and raw ATL03 photon cloud data, a pseudo-waveform is generated for grading the ATL08 terrain estimates. Finally, all the ATL08 terrain estimates are graded based on the peak characteristics of the generated pseudo-waveform. To validate the feasibility of the proposed method, four study areas from the National Ecological Observatory Network (NEON), characterized by various terrain features and forest types were selected. High-accuracy airborne lidar data were used to evaluate the accuracy of graded ICESat-2 terrain estimates. The results demonstrate that the method effectively classified all ATL08 terrain estimates into different accuracy levels and successfully extracted high-accuracy GCPs. The root mean square errors (RMSEs) of the first accuracy level in the four selected study areas were 0.99 m, 0.51 m, 1.88 m, and 0.65 m, representing accuracy improvement of 51.7%, 58.2%, 83.1%, and 68.8%, respectively, compared to the original ATL08 terrain estimates before classifying. Additionally, a comparison with the conventional threshold-based GCP extraction method demonstrated the superior performance of our proposed approach. This study introduces a new approach to extract high-quality elevation control points from ICESat-2 ATL08 data, particularly in forested areas. Full article

Addressing the challenges in online multi-object tracking algorithms under complex scenarios, where the independence among feature extraction, object detection, and data association modules leads to both error accumulation and the difficulty of maintaining visual consistency for occluded objects, we have proposed an end-to-end multi-object tracking method based on hypergraph matching (JDTHM). Initially, a feature extraction and object detection module is introduced to achieve preliminary localization and description of the objects. Subsequently, a deep feature aggregation module is designed to extract temporal information from historical tracklets, amalgamating features from object detection and feature extraction to enhance the consistency between the current frame features and the tracklet features, thus preventing identity swaps and tracklet breaks caused by object detection loss or distortion. Finally, a data association module based on hypergraph matching is constructed, integrating with object detection and feature extraction into a unified network, transforming the data association problem into a hypergraph matching problem between the tracklet hypergraph and the detection hypergraph, thereby achieving end-to-end model optimization. The experimental results demonstrate that this method has yielded favorable qualitative and quantitative analysis results on three multi-object tracking datasets, thereby validating its effectiveness in enhancing the robustness and accuracy of multi-object tracking tasks. Full article

As the combination of panoramic images and laser point clouds becomes more and more widely used as a technique, the accurate determination of external parameters has become essential. However, due to the relative position change of the sensor and the time synchronization error, the automatic and accurate matching of the panoramic image and the point cloud is very challenging. In order to solve this problem, this paper proposes an automatic and accurate registration method for panoramic images and point clouds of urban large scenes based on line features. Firstly, the multi-modal point cloud line feature extraction algorithm is used to extract the edge of the point cloud. Based on the point cloud intensity orthoimage (an orthogonal image based on the point cloud’s intensity values), the edge of the road markings is extracted, and the geometric feature edge is extracted by the 3D voxel method. Using the established virtual projection correspondence for the panoramic image, the panoramic image is projected onto the virtual plane for edge extraction. Secondly, the accurate matching relationship is constructed by using the feature constraint of the direction vector, and the edge features from both sensors are refined and aligned to realize the accurate calculation of the registration parameters. The experimental results show that the proposed method shows excellent registration results in challenging urban scenes. The average registration error is better than 3 pixels, and the root mean square error (RMSE) is less than 1.4 pixels. Compared with the mainstream methods, it has advantages and can promote the further research and application of panoramic images and laser point clouds. Full article

Tourists’ near-field passive perception and identification in mountain areas faces challenges related to long distances, small targets, varied-pose scenarios, facial occlusion, etc. To address this issue, this paper proposes an innovative technical framework based on a face-to-body (F2B) two-step iterative method aimed at enhancing the passive perception and tracking of tourists in complex mountain environments by integrating and coordinating body features with facial features. The F2B technical framework comprises three main components: target feature acquisition, multi-feature coupled re-identification, and target positioning and tracking. Initially, the faces and bodies of tourists are extracted from real-time video streams using the RetinaFace and YOLOX models, respectively. The ArcFace model is then employed to extract the facial features of the target tourists, linking them with the faces detected by RetinaFace. Subsequently, a multi-feature database is constructed using the Hungarian algorithm to facilitate the automatic matching of the face and body of the same tourist. Finally, the Fast-ReID model and a spatial position algorithm are utilized for the re-identification of tourist targets and tracking their dynamic paths. Based on public and actual scene datasets, deployment and testing in the Yimeng Mountain Scenic Area have demonstrated that the accuracy index AP of the F2B model reaches 88.03%, with a recall of 90.28%, achieving an overall identification accuracy of approximately 90% and a false alarm rate of less than 5%. This result significantly improves the accuracy of SOTA facial recognition models in the complex environments of mountainous scenic spots. It effectively addresses the challenges associated with the low identification accuracy of non-cooperative targets in these areas through a ground video sensing network. Furthermore, it offers technical support for spatiotemporal information regarding near-field passive perception and path tracking of tourists in mountain scenic spots and showcasing broad application prospects. Full article