Thanh Binh Ngo
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Thanh Binh Ngo has a total of 43 co-authors. Here are the top co-authors by number of shared publications:
- Andy Nguyen
School of Science, Engineering and Digital Technologies, University of Southern Queensland, Springfield, QLD 4300, Australia
6 shared publications - Long Ngo
Software and Service Development Department, Mobifone Digital Services, Hanoi 100000, Vietnam
5 shared publications - Trung Thi Hoa Trang Nguyen
University of Transport and Communications
5 shared publications - Xuan Chieu Luong
Department of Highway and Traffic Engineering, University of Transport and Communications, Hanoi 100000, Vietnam
5 shared publications - Toan Thanh Dao
University of Transport and Communications
4 shared publications - Anh Vu Phi
Computer Science Department, College of Engineering, Michigan State University, East Lansing, MI 48823, USA
3 shared publications
Latest Publications
- Open Access
- 30 July 2026
- Published by OU Scientific Route in EUREKA: Physics and Engineering
- No. 4,p. 86-100
Implementation of intrinsic and extrinsic calibration of RGB camera and ouster os1 lidar for autonomous mobile robot sensor fusion using ros1
Autonomous mobile robots (AMRs) operating in real-world environments such as warehouses and hospitals require accurate environmental perception for safe navigation. This paper addresses the calibration of heterogeneous sensors – a Red-Green-Blue (RGB) USB camera and an Ouster OS1 Light Detection and Ranging (LiDAR) – and their fusion on a differential-drive AMR running under Robot Operating System version 1 (ROS1) Noetic. The core problem is the absence of pre-recorded extrinsic transforms and the impracticality of live calibration when only offline recorded data are available, rendering standard ROS-based workflows inapplicable. To solve this, a complete three-step pipeline is proposed: 1) offline camera intrinsic calibration from extracted .bag frames using a planar checkerboard method with automated farthest-point sampling for frame diversity; 2) interactive LiDAR-to-camera alignment via axis-mapping selection among 11 rotation candidates followed by fine-tuning with real-time visual feedback; 3) calibration verification by projecting the full LiDAR point cloud onto synchronized camera frames with depth-based colorization. Experiments on a 205-second, 10.5 GB .bag file demonstrate a reprojection error of 0.2748 px (RMS) – well below the 0.5 px excellence threshold – and an average of 7,000–7,500 LiDAR points projected per frame, reaching 12,428 points in optimal scenes. The low error is explained by diversity-based frame selection, which eliminates redundant frames and ensures broad spatial coverage. The pipeline is implemented entirely in Python using standard open-source libraries without GPU, making it suitable for resource-constrained onboard computers in practical robotic deployments.
- 30 July 2026
- Published by OU Scientific Route in EUREKA: Physics and Engineering
- No. 4,p. 86-100
Open AccessImplementation of intrinsic and extrinsic calibration of RGB camera and ouster os1 lidar for autonomous mobile robot sensor fusion using ros1
Autonomous mobile robots (AMRs) operating in real-world environments such as warehouses and hospitals require accurate environmental perception for safe navigation. This paper addresses the calibration of heterogeneous sensors – a Red-Green-Blue (RGB) USB camera and an Ouster OS1 Light Detection and Ranging (LiDAR) – and their fusion on a differential-drive AMR running under Robot Operating System version 1 (ROS1) Noetic. The core problem is the absence of pre-recorded extrinsic transforms and the impracticality of live calibration when only offline recorded data are available, rendering standard ROS-based workflows inapplicable. To solve this, a complete three-step pipeline is proposed: 1) offline camera intrinsic calibration from extracted .bag frames using a planar checkerboard method with automated farthest-point sampling for frame diversity; 2) interactive LiDAR-to-camera alignment via axis-mapping selection among 11 rotation candidates followed by fine-tuning with real-time visual feedback; 3) calibration verification by projecting the full LiDAR point cloud onto synchronized camera frames with depth-based colorization. Experiments on a 205-second, 10.5 GB .bag file demonstrate a reprojection error of 0.2748 px (RMS) – well below the 0.5 px excellence threshold – and an average of 7,000–7,500 LiDAR points projected per frame, reaching 12,428 points in optimal scenes. The low error is explained by diversity-based frame selection, which eliminates redundant frames and ensures broad spatial coverage. The pipeline is implemented entirely in Python using standard open-source libraries without GPU, making it suitable for resource-constrained onboard computers in practical robotic deployments.
- Open Access
TDMA-Based LoRa IoT Architecture with FreeRTOS for Real-Time Multi-Node Bridge Structural Health Monitoring
Structural health monitoring (SHM) systems based on Internet of Things (IoT) technologies have become an effective approach for continuous monitoring of bridge infrastructures. However, many wireless monitoring systems relying on LoRaWAN or contention-based communication suffer from packet collisions, unpredictable latency, and limited scalability when multiple sensing nodes operate simultaneously. To address these limitations, this study proposes a soft real-time LoRa-based IoT architecture for bridge SHM using a time division multiple access (TDMA) communication framework implemented on an embedded real-time platform. The proposed system integrates distributed vibration sensing nodes, a TDMA-enabled LoRa communication layer, an ESP32-based gateway, and a web-based monitoring database for remote visualization and analysis. The architecture leverages FreeRTOS (v10.4.3) for system-level task scheduling, enabling concurrent execution of sensing, communication, and networking processes across the dual-core ESP32-WROOM-32D platform. Experimental results obtained using a laboratory-scale cable-stayed bridge model demonstrate stable multi-node communication with a packet delivery ratio exceeding 95% and predictable TDMA-scheduled transmission cycles with TDMA slots of 100–200 ms under the evaluated operating conditions. The experiments validate end-to-end operation using a representative three-node deployment, while broader scalability is evaluated analytically through the TDMA capacity model and identified as future work for larger physical deployments.
Open AccessTDMA-Based LoRa IoT Architecture with FreeRTOS for Real-Time Multi-Node Bridge Structural Health Monitoring
Structural health monitoring (SHM) systems based on Internet of Things (IoT) technologies have become an effective approach for continuous monitoring of bridge infrastructures. However, many wireless monitoring systems relying on LoRaWAN or contention-based communication suffer from packet collisions, unpredictable latency, and limited scalability when multiple sensing nodes operate simultaneously. To address these limitations, this study proposes a soft real-time LoRa-based IoT architecture for bridge SHM using a time division multiple access (TDMA) communication framework implemented on an embedded real-time platform. The proposed system integrates distributed vibration sensing nodes, a TDMA-enabled LoRa communication layer, an ESP32-based gateway, and a web-based monitoring database for remote visualization and analysis. The architecture leverages FreeRTOS (v10.4.3) for system-level task scheduling, enabling concurrent execution of sensing, communication, and networking processes across the dual-core ESP32-WROOM-32D platform. Experimental results obtained using a laboratory-scale cable-stayed bridge model demonstrate stable multi-node communication with a packet delivery ratio exceeding 95% and predictable TDMA-scheduled transmission cycles with TDMA slots of 100–200 ms under the evaluated operating conditions. The experiments validate end-to-end operation using a representative three-node deployment, while broader scalability is evaluated analytically through the TDMA capacity model and identified as future work for larger physical deployments.
- Open Access
Trustworthy Cyber–Physical Edge-SHM Architecture for Operational Underground Tunnel Crack Monitoring Under Resource-Constrained Conditions
This study presents a low-cost edge-IoT-based structural health monitoring (SHM) architecture for crack monitoring in operational underground tunnels. The system integrates crack-displacement sensing, temperature measurement, ESP32-based edge processing, LoRa/MQTT communication, AES/ECDSA-based data protection, and cloud-based data management. The architecture was validated through a 30-day field campaign at two representative cracks in the Hai Van Tunnel, with measurements acquired at 60 s sampling intervals. The results show that edge-based wavelet–Kalman processing improved measurement stability and reduced high-frequency noise, while the implemented security mechanism introduced only minor latency relative to the monitoring cycle. The two monitored cracks exhibited small micrometer-scale fluctuations associated with temperature variation and showed no cumulative widening trend during the observation period. This study demonstrates the feasibility of campaign-based tunnel crack monitoring using a trustworthy edge-sensing architecture, while longer deployments with more sensing nodes are needed to fully evaluate scalability and operational durability.
Open AccessTrustworthy Cyber–Physical Edge-SHM Architecture for Operational Underground Tunnel Crack Monitoring Under Resource-Constrained Conditions
This study presents a low-cost edge-IoT-based structural health monitoring (SHM) architecture for crack monitoring in operational underground tunnels. The system integrates crack-displacement sensing, temperature measurement, ESP32-based edge processing, LoRa/MQTT communication, AES/ECDSA-based data protection, and cloud-based data management. The architecture was validated through a 30-day field campaign at two representative cracks in the Hai Van Tunnel, with measurements acquired at 60 s sampling intervals. The results show that edge-based wavelet–Kalman processing improved measurement stability and reduced high-frequency noise, while the implemented security mechanism introduced only minor latency relative to the monitoring cycle. The two monitored cracks exhibited small micrometer-scale fluctuations associated with temperature variation and showed no cumulative widening trend during the observation period. This study demonstrates the feasibility of campaign-based tunnel crack monitoring using a trustworthy edge-sensing architecture, while longer deployments with more sensing nodes are needed to fully evaluate scalability and operational durability.
- Top Cited PapersOpen Access
UAV-Based Visual Detection and Tracking of Drowning Victims in Maritime Rescue Operations
Maritime search and rescue (SAR) operations are challenged by vast search areas, poor visibility, and the time-critical nature of victim survival, particularly in dynamic coastal areas. This study presents an intelligent unmanned aerial vehicle (UAV) framework for real-time detection, tracking, and prioritization of people in distress at sea. Unlike existing UAV-based SAR systems that rely on visual sensing or offline human intervention, the proposed framework integrates RGB-thermal multimodal sensing and posture recognition to enhance victim prioritization and survivability estimation. Visual-thermal data support human posture detection, inference of physiological indicators, and autonomous UAV navigation. Metadata are transmitted to a ground control station to enable adaptive altitude control, trajectory rejoining, and multi-target prioritization. Field-inspired experiments in Quang Ninh Province, Vietnam demonstrated robust real-time performance, achieving 23 FPS with detection accuracy up to 84% for swimming subjects and over 50% for drowning postures. These findings demonstrate that Edge-AI-enabled UAVs can serve as a practical and efficient solution for maritime SAR, reducing response times and improving mission outcomes.
Top Cited PapersOpen AccessUAV-Based Visual Detection and Tracking of Drowning Victims in Maritime Rescue Operations
Maritime search and rescue (SAR) operations are challenged by vast search areas, poor visibility, and the time-critical nature of victim survival, particularly in dynamic coastal areas. This study presents an intelligent unmanned aerial vehicle (UAV) framework for real-time detection, tracking, and prioritization of people in distress at sea. Unlike existing UAV-based SAR systems that rely on visual sensing or offline human intervention, the proposed framework integrates RGB-thermal multimodal sensing and posture recognition to enhance victim prioritization and survivability estimation. Visual-thermal data support human posture detection, inference of physiological indicators, and autonomous UAV navigation. Metadata are transmitted to a ground control station to enable adaptive altitude control, trajectory rejoining, and multi-target prioritization. Field-inspired experiments in Quang Ninh Province, Vietnam demonstrated robust real-time performance, achieving 23 FPS with detection accuracy up to 84% for swimming subjects and over 50% for drowning postures. These findings demonstrate that Edge-AI-enabled UAVs can serve as a practical and efficient solution for maritime SAR, reducing response times and improving mission outcomes.
- Open Access
- 29 September 2025
- Published by MDPI AG in Journal of Sensor and Actuator Networks
- Vol. 14 (5) , 98
Drone Imaging and Sensors for Situational Awareness in Hazardous Environments: A Systematic Review
Situation awareness is essential for ensuring safety in hazardous environments, where timely and accurate information is critical for decision-making. Unmanned Aerial Vehicles (UAVs) have emerged as valuable tools in enhancing situation awareness by providing real-time data and monitoring capabilities in high-risk areas. This study explores the integration of advanced technologies, focusing on imaging and sensor technologies such as thermal, spectral, and multispectral cameras, deployed in critical zones. By merging these technologies into UAV platforms, responders gain access to essential real-time information while reducing human exposure to hazardous conditions. This study presents case studies and practical applications, highlighting the effectiveness of these technologies in a range of hazardous situations.
- 29 September 2025
- Published by MDPI AG in Journal of Sensor and Actuator Networks
- Vol. 14 (5) , 98
Open AccessDrone Imaging and Sensors for Situational Awareness in Hazardous Environments: A Systematic Review
Situation awareness is essential for ensuring safety in hazardous environments, where timely and accurate information is critical for decision-making. Unmanned Aerial Vehicles (UAVs) have emerged as valuable tools in enhancing situation awareness by providing real-time data and monitoring capabilities in high-risk areas. This study explores the integration of advanced technologies, focusing on imaging and sensor technologies such as thermal, spectral, and multispectral cameras, deployed in critical zones. By merging these technologies into UAV platforms, responders gain access to essential real-time information while reducing human exposure to hazardous conditions. This study presents case studies and practical applications, highlighting the effectiveness of these technologies in a range of hazardous situations.
- Open Access
C2L3-Fusion: An Integrated 3D Object Detection Method for Autonomous Vehicles
Accurate 3D object detection is crucial for autonomous vehicles (AVs) to navigate safely in complex environments. This paper introduces a novel fusion framework that integrates Camera image-based 2D object detection using YOLOv8 and LiDAR data-based 3D object detection using PointPillars, hence named C2L3-Fusion. Unlike conventional fusion approaches, which often struggle with feature misalignment, C2L3-Fusion enhances spatial consistency and multi-level feature aggregation, significantly improving detection accuracy. Our method outperforms state-of-the-art approaches such as YoPi-CLOCs Fusion Network, standalone YOLOv8, and standalone PointPillars, achieving mean Average Precision (mAP) scores of 89.91% (easy), 79.26% (moderate), and 78.01% (hard) on the KITTI dataset. Successfully implemented on the Nvidia Jetson AGX Xavier embedded platform, C2L3-Fusion maintains real-time performance while enhancing robustness, making it highly suitable for self-driving vehicles. This paper details the methodology, mathematical formulations, algorithmic advancements, and real-world testing of C2L3-Fusion, offering a comprehensive solution for 3D object detection in autonomous navigation.
Open AccessC2L3-Fusion: An Integrated 3D Object Detection Method for Autonomous Vehicles
Accurate 3D object detection is crucial for autonomous vehicles (AVs) to navigate safely in complex environments. This paper introduces a novel fusion framework that integrates Camera image-based 2D object detection using YOLOv8 and LiDAR data-based 3D object detection using PointPillars, hence named C2L3-Fusion. Unlike conventional fusion approaches, which often struggle with feature misalignment, C2L3-Fusion enhances spatial consistency and multi-level feature aggregation, significantly improving detection accuracy. Our method outperforms state-of-the-art approaches such as YoPi-CLOCs Fusion Network, standalone YOLOv8, and standalone PointPillars, achieving mean Average Precision (mAP) scores of 89.91% (easy), 79.26% (moderate), and 78.01% (hard) on the KITTI dataset. Successfully implemented on the Nvidia Jetson AGX Xavier embedded platform, C2L3-Fusion maintains real-time performance while enhancing robustness, making it highly suitable for self-driving vehicles. This paper details the methodology, mathematical formulations, algorithmic advancements, and real-world testing of C2L3-Fusion, offering a comprehensive solution for 3D object detection in autonomous navigation.
- Open Access
- 19 March 2025
- Published by University of Transport and Communications in Transport and Communications Science Journal
- Vol. 76 (4) , 489-503
Crack detection on concrete surfaces using the YOLOv8 quantization model
The condition of concrete structures’ surfaces, most importantly the condition of cracks on the surface and their development over time, is an important and common criterion used to diagnose health conditions and determine their service life. Rapid collection, identification, and monitoring of concrete structures’ surfaces to assess the condition of bridge structures requires a fast-acting system that can meet the actual speed of the inspection process. This article introduces a YOLOv8 quantization model for rapid crack detection in concrete structures, leveraging GPU acceleration suitable for real-time video analysis during bridge inspections. The method incorporates histogram equalization and image enhancement to mitigate lighting issues and improve crack visibility. INT8 quantization reduces model size and accelerates processing while maintaining accuracy through dataset calibration. Converted to TensorRT and integrated into the inference pipeline for optimized GPU and memory management, the YOLOv8 quantization model achieves at least 30 FPS using full HD video footage. Field tests with NVIDIA GPUs demonstrated a 5x reduction in processing time, a 5x FPS increase, and a 6x improvement in GPU utilization, all while maintaining similar RAM usage. The quantum YOLOv8 model is optimized for NVIDIA GPUs, achieving a balance between accuracy and processing speed, allowing workers to analyze full HD videos in real-time at a rate of at least 30 FPS during field inspections
- 19 March 2025
- Published by University of Transport and Communications in Transport and Communications Science Journal
- Vol. 76 (4) , 489-503
Open AccessCrack detection on concrete surfaces using the YOLOv8 quantization model
The condition of concrete structures’ surfaces, most importantly the condition of cracks on the surface and their development over time, is an important and common criterion used to diagnose health conditions and determine their service life. Rapid collection, identification, and monitoring of concrete structures’ surfaces to assess the condition of bridge structures requires a fast-acting system that can meet the actual speed of the inspection process. This article introduces a YOLOv8 quantization model for rapid crack detection in concrete structures, leveraging GPU acceleration suitable for real-time video analysis during bridge inspections. The method incorporates histogram equalization and image enhancement to mitigate lighting issues and improve crack visibility. INT8 quantization reduces model size and accelerates processing while maintaining accuracy through dataset calibration. Converted to TensorRT and integrated into the inference pipeline for optimized GPU and memory management, the YOLOv8 quantization model achieves at least 30 FPS using full HD video footage. Field tests with NVIDIA GPUs demonstrated a 5x reduction in processing time, a 5x FPS increase, and a 6x improvement in GPU utilization, all while maintaining similar RAM usage. The quantum YOLOv8 model is optimized for NVIDIA GPUs, achieving a balance between accuracy and processing speed, allowing workers to analyze full HD videos in real-time at a rate of at least 30 FPS during field inspections
- Open Access
- 15 January 2025
- Published by University of Transport and Communications in Transport and Communications Science Journal
- Vol. 76 (1) , 64-79
Real-time multi-sensor fusion for object detection and localization in self-driving cars: A Carla simulation
Research on integrating camera and LiDAR in self-driving car systems has important scientific significance in the context of developing 4.0 technology and applying artificial intelligence. The research contributes to improving the accuracy in recognizing and locating objects in complex environments. This is an important foundation for further research on optimizing response time and improving the safety of self-driving systems. This study proposes a real-time multi-sensor data fusion method, termed "Multi-Layer Fusion," for object detection and localization in autonomous vehicles. The fusion process leverages pixel-level and feature-level integration, ensuring seamless data synchronization and robust performance under adverse conditions. Experiments conducted on the CARLA simulator. The results show that the method significantly improves environmental perception and object localization, achieving a mean detection accuracy of 95% and a mean distance error of 0.54 meters across diverse conditions, with real-time performance at 30 FPS. These results demonstrate its robustness in both ideal and adverse scenarios
- 15 January 2025
- Published by University of Transport and Communications in Transport and Communications Science Journal
- Vol. 76 (1) , 64-79
Open AccessReal-time multi-sensor fusion for object detection and localization in self-driving cars: A Carla simulation
Research on integrating camera and LiDAR in self-driving car systems has important scientific significance in the context of developing 4.0 technology and applying artificial intelligence. The research contributes to improving the accuracy in recognizing and locating objects in complex environments. This is an important foundation for further research on optimizing response time and improving the safety of self-driving systems. This study proposes a real-time multi-sensor data fusion method, termed "Multi-Layer Fusion," for object detection and localization in autonomous vehicles. The fusion process leverages pixel-level and feature-level integration, ensuring seamless data synchronization and robust performance under adverse conditions. Experiments conducted on the CARLA simulator. The results show that the method significantly improves environmental perception and object localization, achieving a mean detection accuracy of 95% and a mean distance error of 0.54 meters across diverse conditions, with real-time performance at 30 FPS. These results demonstrate its robustness in both ideal and adverse scenarios
- Open Access
Optimising Concrete Crack Detection: A Study of Transfer Learning with Application on Nvidia Jetson Nano
The use of Artificial Intelligence (AI) to detect defects such as concrete cracks in civil and transport infrastructure has the potential to make inspections less expensive, quicker, safer and more objective by reducing the need for on-site human labour. One deployment scenario involves using a drone to carry an embedded device and camera, with the device making localised predictions at the edge about the existence of defects using a trained convolutional neural network (CNN) for image classification. In this paper, we trained six CNNs, namely Resnet18, Resnet50, GoogLeNet, MobileNetV2, MobileNetV3-Small and MobileNetV3-Large, using transfer learning technology to classify images of concrete structures as containing a crack or not. To enhance the model’s robustness, the original dataset, comprising 3000 images of concrete structures, was augmented using salt and pepper noise, as well as motion blur, separately. The results show that Resnet50 generally provides the highest validation accuracy (96% with the original dataset and a batch size of 16) and the highest validation F1-score (95% with the original dataset and a batch size of 16). The trained model was then deployed on an Nvidia Jetson Nano device for real-time inference, demonstrating its capability to accurately detect cracks in both laboratory and field settings. This study highlights the potential of using transfer learning on Edge AI devices for Structural Health Monitoring, providing a cost-effective and efficient solution for automated crack detection in concrete structures.
Open AccessOptimising Concrete Crack Detection: A Study of Transfer Learning with Application on Nvidia Jetson Nano
The use of Artificial Intelligence (AI) to detect defects such as concrete cracks in civil and transport infrastructure has the potential to make inspections less expensive, quicker, safer and more objective by reducing the need for on-site human labour. One deployment scenario involves using a drone to carry an embedded device and camera, with the device making localised predictions at the edge about the existence of defects using a trained convolutional neural network (CNN) for image classification. In this paper, we trained six CNNs, namely Resnet18, Resnet50, GoogLeNet, MobileNetV2, MobileNetV3-Small and MobileNetV3-Large, using transfer learning technology to classify images of concrete structures as containing a crack or not. To enhance the model’s robustness, the original dataset, comprising 3000 images of concrete structures, was augmented using salt and pepper noise, as well as motion blur, separately. The results show that Resnet50 generally provides the highest validation accuracy (96% with the original dataset and a batch size of 16) and the highest validation F1-score (95% with the original dataset and a batch size of 16). The trained model was then deployed on an Nvidia Jetson Nano device for real-time inference, demonstrating its capability to accurately detect cracks in both laboratory and field settings. This study highlights the potential of using transfer learning on Edge AI devices for Structural Health Monitoring, providing a cost-effective and efficient solution for automated crack detection in concrete structures.