BY-SLAM: Dynamic Visual SLAM System Based on BEBLID and Semantic Information Extraction.

Existing dynamic scene visual SLAM methods indiscriminately remove feature points on dynamic target bounding boxes when using object detection algorithms to detect prior targets, resulting in the incorrect removal of some valid static points and affecting the localization accuracy of the algorithm. This paper proposes a visual SDT-SLAM algorithm, which uses YOLOv11 to determine the prior target information of each frame of the image and designs a short-term dynamic object tracking algorithm to track targets and classify feature points in real time for different categories of bounding boxes. In the state estimation process, a depth difference estimation algorithm is used to recover some erroneously deleted static feature points, and an epipolar constraint algorithm is used to filter dynamic feature points from the scene in a fine-grained manner. This method precisely distinguishes the dynamic and static feature points in dynamic scenes. Compared with other SLAM systems, this system reasonably utilizes depth information to restore static feature points in dynamic scenes, and based on semantic information and epipolar constraints, it removes incorrect matching points, significantly preserving and fully utilizing the static feature points in dynamic scenes. This further improves the trajectory estimation accuracy of the SLAM algorithm in dynamic scenes. Without relying on computationally intensive segmentation methods, it achieves a new balance between real-time performance and accuracy. The SDT-SLAM algorithm was validated on the TUM and Bonn public datasets and in the real world. The results show that SDT-SLAM has advantages in speed and accuracy for localization in highly dynamic environments compared to other advanced algorithms.

In semantic network-based dynamic scene visual SLAM methods, the masks generated by semantic networks often over-cover dynamic objects, resulting in parts of the static background being mistakenly included in the masks. Furthermore, the system fails to identify passively moving potential dynamic objects outside the masks, which impacts the pose estimation effect. To address these challenges, this paper proposes a mask-filtering method that integrates image depth information. By utilizing the consistency of object depth, this approach effectively eliminates background regions mistakenly included in the mask. Subsequently, we developed a regional dynamic point rejection strategy and proposed an anomaly detection method based on an adaptive Gaussian model. By constructing an adaptive Gaussian model for dynamic points inside the mask, setting dynamic thresholds, and updating the model parameters, this method can detect potential dynamic points outside the mask. It effectively reduces the impact of potential dynamic objects on the system’s pose estimation. Finally, this paper presents a dynamic Gaussian SLAM system based on the ORB-SLAM3 framework, named DG-SLAM. This system is used for pose estimation and dense point cloud construction, improving localization accuracy in dynamic scenes. To validate the performance of the DG-SLAM system, we conducted tests on the TUM RGBD dataset and in real-world scenarios, comparing it with ORB-SLAM3, DynaSLAM, and SG-SLAM. Compared to ORB-SLAM3, DG-SLAM achieves an average improvement of 94.54% and a minimum improvement of 89.34% in positioning accuracy on the TUM and Bonn datasets. Experimental results show that DG-SLAM can effectively detect and eliminate the interference of potential dynamic feature points in the scene, achieving good localization accuracy and strong robustness in various dynamic environments.

Simultaneous localization and map building (SLAM) is crucial in autonomous robot navigation. However, existing SLAM systems generally assume a static environment, which makes it difficult to cope with the interference caused by moving objects in dynamic scenes, affecting the system’s localization accuracy and robustness. To address this challenge, this paper proposes YKD-SLAM, a visual SLAM system for indoor dynamic environments, which is based on the ORB-SLAM2 framework and incorporates YOLOv8 target detection, RCF-KMeans (Region-ConstrainedFastK-Means), and epipolar geometric constraints to realize the accurate rejection of dynamic feature points and improve the localization performance in dynamic environments. YKD-SLAM first uses YOLOv8 to detect dynamic objects in the scene, generates a detection frame, optimizes the depth map through open operations, and performs multi-region segmentation of the region within the detection frame by combining RCF-KMeans. Subsequently, through the dynamic feature point rejection strategy based on epipolar geometric constraints, different regions in the detection frame are discriminated into dynamic and static regions, and the feature points in the dynamic region are rejected to improve the localization accuracy and robustness of the system in dynamic environments. The experimental results show that YKD-SLAM performs well in several dynamic scenes in the TUMRGB-D dataset. Compared with ORB-SLAM2, its ATE is reduced by 98.37%; compared with DynaSLAM, the system operation efficiency is improved by 95.35%. In addition, practical experiments conducted in indoor dynamic scenes further validate its potential in real applications.

This paper proposes a semantic information extraction technology model system based on BERT, aiming to solve the problem of semantic understanding and information extraction in online public opinion analysis. The system design includes data collection, preprocessing, semantic information extraction and result display modules, using the BERT model to capture the deep semantic relationship of the text and enhance the ability to understand complex semantics and contextual information. The data collection module covers multiple sources such as social media, news comments and forum posts, and ensures the high quality and diversity of data through real-time crawling and cleaning. The semantic information extraction module is based on the multi-head attention mechanism of BERT, which can effectively capture the emotional tendency, topic information and potential intention in the public opinion text. The experiment verifies the system performance based on 1 million online public opinion data. The results show that in the semantic information extraction task, the accuracy of the BERT-based model is improved by 22.4%, the recall rate is improved by 19.8%, and the F1 value reaches 91.7<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">%</sup>, which is significantly better than traditional methods such as TF-IDF and Word2Vec. The sentiment classification module has an accuracy of 93.1%, achieving accurate recognition of positive, negative and neutral emotions. The processing time of a single piece of public opinion text is reduced by an average of 25.6%, indicating the high efficiency of the system in large-scale data processing. The simulation test verifies the system's stable operation capability in the scenario of 100,000 new public opinion data per day, with a delay of less than 1 second, which can meet the needs of real-time public opinion analysis.

The autonomous navigation capability of USVs in complex water environments relies on Simultaneous Localization and Mapping (SLAM) technology. However, water surface perturbations and reflections generate numerous dynamic invalid feature points, causing low localization accuracy and poor robustness in VSLAM. To this end, we propose a monocular VSLAM algorithm considering dynamic invalid features of the water surface for USV, named WS-SLAM. Firstly, the ORBSLAM3 framework is improved by adding the PP-LiteSeg semantic segmentation thread to distinguish between dynamic invalid feature points of the water surface and feature points on shore. Secondly, semantic information is used to generate a mask to cull dynamic invalid feature points of the water surface, reducing the impact of water current perturbation on visual localization. Then, further culling of potentially residual dynamic feature points on shore is performed in conjunction with the epipolar geometric constraint. Finally, the camera position is estimated using only static feature points. The experimental results based on the public USVInland dataset show that WSSLAM improves the positioning accuracy by an average of 43.47% in straight line sailing compared with ORBSLAM3. It can effectively inhibit the position drift caused by the dynamic invalid features of the water surface and has higher positioning accuracy and stronger robustness.

Visual simultaneous localization and mapping (SLAM) algorithms in dynamic scenes will apply the moving feature points to the camera pose’s calculation, which will cause the continuous accumulation of errors. As a target-detection tool, mask R-CNN, which is often used in combination with the former, due to the limited training datasets, easily results in the semantic mask being incomplete and deformed, which will increase the error. In order to solve the above problems, we propose in this paper a visual SLAM algorithm based on an adaptive mask dilation strategy and the dynamic probability of the feature points, named MDP-SLAM. Firstly, we use the mask R-CNN target-detection algorithm to obtain the initial mask of the dynamic target. On this basis, an adaptive mask-dilation algorithm is used to obtain a mask that can completely cover the dynamic target and part of the surrounding scene. Then, we use the K-means clustering algorithm to segment the depth image information in the mask coverage area into absolute dynamic regions and relative dynamic regions. Combined with the epipolar constraint and the semantic constraint, the dynamic probability of the feature points is calculated, and then, the highly dynamic possible feature points are removed to solve an accurate final pose of the camera. Finally, the method is tested on the TUM RGB-D dataset. The results show that the MDP-SLAM algorithm proposed in this paper can effectively improve the accuracy of attitude estimation and has high accuracy and robustness in dynamic indoor scenes.

Current visual simultaneous localization and mapping (SLAM) systems have demonstrated commendable efficacy in static environments. However, the presence of dynamic objects in real-world settings frequently leads to system discrepancies, significantly impairing the accuracy and robustness of SLAM systems. Conventional visual SLAM approaches typically utilize epipolar constraints to mitigate the impact of outliers; nevertheless, they encounter limitations when confronted with a substantial number of dynamic or planar moving objects. To tackle these issues, this paper introduces a novel dynamic visual SLAM system, termed DEG-SLAM. Initially, the system employs the YOLOv5 object detection network to identify dynamic objects, subsequently relaying the semantic information to the tracking module. During the tracking phase, both semantic information and epipolar constraints are leveraged to filter out dynamic feature points. To address the challenges posed by the malfunctioning of epipolar constraints in degenerate scenes, DEG-SLAM incorporates a degenerate constraint mechanism aimed at further eliminating dynamic feature points. Furthermore, a reprojection constraint has been introduced to enhance the filtering of absent dynamic feature points that lie outside the detection boxes. Experimental findings reveal that DEG-SLAM significantly improves accuracy and robustness when compared to traditional ORB-SLAM3 in dynamic environments. The performance benefits of DEG-SLAM are particularly evident in degenerate scenarios, thereby affirming its practicality and reliability in complex settings.

To address the challenges of visual SLAM algorithms in unmanned surface vehicles (USVs) during nearshore navigation or docking, where dynamic feature points degrade localization accuracy and dynamic objects impede static dense mapping, this study proposes an improved visual SLAM algorithm that removes dynamic feature points. Building upon the ORB-SLAM3 framework, the improved SLAM algorithm integrates a shore segmentation module and a dynamic region elimination module, while enabling static dense point cloud mapping. The system first implements shore segmentation based on Otsu’s method to generate masks covering water and sky regions, ensuring the SLAM system avoids extracting interfering feature points from these areas. Secondly, the deep learning network YOLOv8n-seg is employed to detect priori dynamic objects, with the motion consistency check method to identify non-priori dynamic feature points, collectively removing dynamic feature points. Additionally, the ELAS algorithm computes disparity maps, integrating depth information and dynamic object information to construct a static dense map. Experimental results demonstrate that, compared to the original ORB-SLAM3, the improved SLAM algorithm achieves superior localization accuracy in dynamic nearshore environments, significantly reduces the impact of dynamic objects on pose estimation, and successfully constructs ghosting-free static dense point cloud maps.

The accuracy of camera pose estimation and map construction is an important factor on availability of path planning, which might be reduced due to the interference from dynamic objects. This paper proposes a dense dynamic SLAM system based on motion element filtering, which realizes the construction of background environment map and the restoration the motion trajectory of dynamic elements. Firstly, after detecting the moving objects in image frames by YOLOv5x algorithm, the system matches ORB feature points to find out the consistency in all objects. Then, based on the results of object detection and semantic segmentation, the dynamic and static feature points in the images are distinguished. After that, the system uses all the static feature points to estimate the pose of the sensor, and constructing a dense background OctoMap. Simultaneously, after estimating the relative pose of each moving object, a global dynamic visual odometry is built by restoring the motion trajectory and real-time inserting the corresponding 3D models in a point cloud space. The experimental results show that the proposed method has better localization effect than ORB-SLAM2 in the dynamic environment. And the system can densely recover background map matching around environment, and globally locate and track the dynamic elements, which retains the dynamic information in the scene.

Most mainstream visual-inertial SLAM systems rely on the point feature method of front-end images for motion estimation and localization. However, the SLAM system based on the point feature method, in the case of weakly textured environments, may not be able to estimate the motion due to the lack of point features. For this reason, the introduction of line features has received extensive attention. Line features in the environment can improve the localization accuracy of SLAM, but the extraction and matching of line features are relatively time-consuming and cannot satisfy real-time requirements. Based on the EDLines algorithm, we propose a fast visual-inertial odometry that fuses point and line features. The improved EDLines algorithm is used to process line features. By adjusting the internal parameters of the algorithm, the extraction time of line features is decreased while the extraction effect is the same. And the line features screening strategy is proposed to filter out invalid short line segments and merge adjacent line segments. We use the more effective line features after screening for graph matching and back-end optimization, which reduces the calculation amount of line features matching and ensures the accuracy. The proposed algorithm has been tested on the public datasets, and the comparison results with PL-VIO and VINS-Mono show that the algorithm in this paper improved real-time performance and localization accuracy.

Most existing vision‐based simultaneous localization and mapping systems and their variants still assume that the observation is absolutely static and cannot work well in dynamic environments. In this paper, we propose a direct geometrically constrained SLAM method based on target detection and depth image segmentation, named YGDD‐SLAM. The YGDD‐SLAM system can work robustly, accurately, and continuously in highly dynamic environments. The method first acquires static and potential dynamic feature points in the current frame through a target detection network. Then, dynamic targets are identified by combining the geometric change relationship between static and potential dynamic feature points between adjacent frames. To improve the accuracy of the dynamic judgment, the motion probability of the potential dynamic target in the past few frames is also used for judgment. Subsequently, the dynamic object regions at the pixel level are segmented out based on the double‐peak feature of the gray‐scale histogram of the dynamic target region in the depth image, which ultimately achieves the accurate deletion of all dynamic features points. Meanwhile, we validate YGDD‐SLAM on TUM data set and Bonn data set and prove that it significantly improves the localization accuracy and system stability in different types of dynamic environments.

Visual SLAM is capable of achieving the localization and mapping functions for intelligent vehicles. However, in practical applications, the presence of complex environments significantly affects localization accuracy and real-time performance. To address these challenges, this paper proposes a dynamic scene visual SLAM method that fuses image segmentation and feature following. This method utilizes an object detection model to quickly identify dynamic regions within image, actively removes unstable dynamic feature points. By employing a clustering algorithm to partition map points and efficiently extract feature-rich regions, this approach avoids interference from dynamic objects. Simultaneously, it utilizes an optical flow algorithm to continuously track static feature-rich regions in subsequent images, mitigating the influence of dynamic pixels and achieving efficient feature point extraction. Experimental results demonstrate that in dynamic environments, the proposed method improves localization accuracy by 70% and reduces feature extraction time per frame by 45% compared to the ORB-SLAM2 system. In most cases, our method also outperforms DS-SLAM, which proves that the proposed method can effectively improve the localization accuracy and the real-time performance of feature extraction in dynamic scenes for visual SLAM.

Generally speaking, visual SLAM systems just based on point features have poor robustness in low-textured scenes, which limits their application. To improve the robustness and accuracy of simultaneous location and mapping in various environments, a multi-scene adaptive visual SLAM system based on RGB-D camera is proposed. We not only use point features in our system, but also introduce line features that are abundant indoors. They are less sensitive to lighting variation and more stable than point features in low-textured scenes. The SLAM system combining point features and line features contains several parts: visual odometry, local mapping, loop closing, full BA and mapping. The experimental results on datasets demonstrate that the performance of our proposed SLAM system is better than state-of-the-art point-based method in various indoor scenes.

In recent years, dynamic visual SLAM techniques have been widely used in autonomous navigation, augmented reality, and virtual reality. However, the increasing demand for computational resources by SLAM techniques limits its application on resource-constrained mobile devices. MEC technology combined with 5G ultra-dense networks enables complex computational tasks in visual SLAM systems to be offloaded to edge computing servers, thus breaking the resource constraints of terminals and meeting real-time computing requirements. This paper firstly introduces the research results in the field of visual SLAM in detail through three categories: static SLAM, dynamic SLAM, and SLAM techniques combined with deep learning. Secondly, the three major parts of the technology comparison between mobile edge computing and mobile cloud computing, 5G ultra-dense networking technology, and MEC and UDN integration technology are introduced to sort out the basic technologies related to the application of 5G ultra-dense network to offload complex computing tasks from visual SLAM systems to edge computing servers.

Simultaneous localization and mapping (SLAM) is essential for mobile robotics, enabling autonomous navigation in dynamic, unstructured outdoor environments without relying on external positioning systems. These environments pose significant challenges due to variable lighting, weather conditions, and complex terrain. Visual‐Inertial SLAM has emerged as a promising solution for robust localization under such conditions. This paper benchmarks several open‐source visual‐Inertial SLAM systems, including traditional methods (ORB‐SLAM3, VINS‐Fusion, OpenVINS, Kimera, and SVO Pro) and learning‐based approaches (HFNet‐SLAM, AirSLAM), to evaluate their performance in unstructured natural outdoor settings. We focus on the impact of loop closing on localization accuracy and computational demands, providing a comprehensive analysis of these systems' effectiveness in real‐world environments and especially their application to embedded systems in outdoor robotics. Our contributions further include an assessment of varying frame rates on localization accuracy and computational load. The findings highlight the importance of loop closing in improving localization accuracy while managing computational resources efficiently, offering valuable insights for optimizing Visual‐Inertial SLAM systems for practical outdoor applications in mobile robotics. The data set and the benchmark code are available under https://github.com/iis-esslingen/vi-slam_lc_benchmark .