Real-time visual SLAM based YOLO-Fastest for dynamic scenes

Measurement Science and Technology, Год журнала: 2024, Номер 35(5), С. 056305 - 056305

Опубликована: Фев. 5, 2024

Abstract Within the realm of autonomous robotic navigation, simultaneous localization and mapping (SLAM) serves as a critical perception technology, drawing heightened attention in contemporary research. The traditional SLAM systems perform well static environments, but real physical world, dynamic objects can destroy geometric constraints system, further limiting its practical application world. In this paper, robust RGB-D system is proposed to expand number points scene by combining with YOLO-Fastest ensure effectiveness model construction, then based on that, new thresholding designed differentiate features objection bounding box, which takes advantage double polyline residuals after reprojection filter feature points. addition, two Gaussian models are constructed segment moving box depth image achieve effect similar instance segmentation under premise ensuring computational speed. experiments conducted sequences provided TUM dataset evaluate performance method, results show that root mean squared error metric absolute trajectory algorithm paper has at least 80% improvement compared ORB-SLAM2. Higher robustness environments both high low DS-SLAM Dynaslam, effectively provide intelligent navigation for mobile robots.

Язык: Английский

---

Data integration using deep learning and real-time locating system (RTLS) for automated construction progress monitoring and reporting

Automation in Construction, Год журнала: 2024, Номер 168, С. 105778 - 105778

Опубликована: Сен. 23, 2024

Язык: Английский

---

Centerline-based registration for shield tunnel 3D reconstruction using spinning mid-range LiDAR point cloud and multi-cameras

Automation in Construction, Год журнала: 2025, Номер 171, С. 105950 - 105950

Опубликована: Янв. 10, 2025

Язык: Английский

---

Understanding digital construction management

Elsevier eBooks, Год журнала: 2025, Номер unknown, С. 27 - 38

Опубликована: Янв. 1, 2025

Язык: Английский

---

SLAM-IMU coupling-based construction personnel positioning in invalid scenario of satellite and aerial triangulation

Engineering Construction & Architectural Management, Год журнала: 2025, Номер unknown

Опубликована: Фев. 19, 2025

Purpose To help supervisors understand the positions of workers in real-time, provide safety guidance for and reduce occurrence accidents. This study proposes a real-time positioning algorithm based on multi-source information coupling, aiming to solve problem workers’ autonomous signal-blind areas. Design/methodology/approach The proposed utilizes visual SLAM IMU sensors perceive environment, construct three-dimensional images, improve accuracy corner point matching, pre-integrate raw data, adopt tightly coupled method couple inertial navigation thereby establishing binocular coupling model. Findings technology shows good effect calculation speed indoor sites, has adaptability different building construction scenarios, error can be controlled within 3%. Originality/value successful effectively alleviates inaccurate caused by signal blind areas existing management system, helps protect lives site improves efficiency supervisors.

Язык: Английский

---

Pose Graph Relocalization with Deep Object Detection and BIM-Supported Object Landmark Dictionary

Journal of Computing in Civil Engineering, Год журнала: 2023, Номер 37(5)

Опубликована: Май 19, 2023

Indoor localization is a prerequisite for autonomous robot applications in the construction industry. However, traditional techniques rely on low-level features and do not exploit construction-related semantics. They also are sensitive to environmental factors such as illumination reflection rate, therefore suffer from unexpected drifts failures. This study proposes pose graph relocalization framework that utilizes object-level landmarks enhance visual system. The proposed builds an object landmark dictionary Building Information Model (BIM) prior knowledge. Then multimodal deep neural network (DNN) realize 3D detection real time, followed by instance-level association with false-positive rejection, relative estimation outlier removal. Finally, keyframe-based optimization performed rectify of localization. was validated using mobile platform red-green-blue-depth (RGB-D) inertial sensors, test scene indoor office environment furnishing elements. model achieved 62.9% mean average precision (mAP). technique reduced translational 64.67% rotational 41.59% compared visual–inertial odometry.

Язык: Английский

---

Unsupervised video object segmentation for enhanced SLAM-based localization in dynamic construction environments

Automation in Construction, Год журнала: 2023, Номер 158, С. 105235 - 105235

Опубликована: Дек. 11, 2023

Язык: Английский

---

Hierarchical Clustering-Based Image Retrieval for Indoor Visual Localization

Electronics, Год журнала: 2022, Номер 11(21), С. 3609 - 3609

Опубликована: Ноя. 4, 2022

Visual localization is employed for indoor navigation and embedded in various applications, such as augmented reality mixed reality. Image retrieval geometrical measurement are the primary steps visual localization, key to improving efficiency reduce time consumption of image retrieval. Therefore, a hierarchical clustering-based image-retrieval method proposed hierarchically organize an off-line database, resulting control within reasonable range. The database organized by two stages: scene-level clustering sub-scene-level clustering. In clustering, improved cumulative sum algorithm detect change points then group images global features. On basis feature tracking-based introduced further into clusters. An with backtracking mechanism designed applied localization. addition, weighted KNN-based presented, estimated query position solved Armijo–Goldstein algorithm. Experimental results indicate that running does not linearly increase size databases, which beneficial efficiency.

Язык: Английский

---

Two-stage RFID approach for localizing objects in smart homes based on gradient boosted decision trees with under- and over-sampling

Journal of Reliable Intelligent Environments, Год журнала: 2023, Номер 10(1), С. 45 - 54

Опубликована: Янв. 12, 2023

Язык: Английский

---

Seamless Fusion: Multi-Modal Localization for First Responders in Challenging Environments

Sensors, Год журнала: 2024, Номер 24(9), С. 2864 - 2864

Опубликована: Апрель 30, 2024

In dynamic and unpredictable environments, the precise localization of first responders rescuers is crucial for effective incident response. This paper introduces a novel approach leveraging three complementary modalities: visual-based, Galileo-based, inertial-based. Each modality contributes uniquely to final Fusion tool, facilitating seamless indoor outdoor localization, offering robust accurate solution without reliance on pre-existing infrastructure, essential maintaining responder safety optimizing operational effectiveness. The visual-based method utilizes an RGB camera coupled with modified implementation ORB-SLAM2 method, enabling operation or prior area scanning. Galileo-based employs lightweight prototype equipped high-accuracy GNSS receiver board, tailored meet specific needs responders. inertial-based sensor fusion, primarily smartphone inertial measurement units, predict adjust responders’ positions incrementally, compensating GPS signal attenuation indoors. A comprehensive validation test involving various environmental conditions was carried out demonstrate efficacy proposed fused tool. Our results show that our always provides location regardless (indoors, outdoors, etc.), overall mean error 1.73 m.

Язык: Английский

---