Article
Engineering, Electrical & Electronic
I-Chen Sang, William R. Norris
Summary: Underwater pipeline inspection is important but using ROVs is costly and fixed leak detection sensors have limited precision. AUVs can significantly reduce costs but face challenges such as unstable currents, low visibility, and loss of GPS signal. This article presents a navigation system for an AUV that incorporates vision and sonar sensors to find and navigate toward the pipeline and inspect it with a resolution of 3m.
IEEE SENSORS JOURNAL
(2023)
Article
Engineering, Ocean
Hongwei Zhang, Shitong Zhang, Yanhui Wang, Yuhong Liu, Yanan Yang, Tian Zhou, Hongyu Bian
Summary: This study developed a scheme using AUV equipped with MBES and FLS for automatic inspection of submarine pipelines, with sea trial results verifying its effectiveness. The adoption of VBS system enhances navigation efficiency, and an improved Otsu algorithm is proposed to improve obstacle avoidance.
APPLIED OCEAN RESEARCH
(2021)
Article
Engineering, Marine
Tianlei Wang, Fei Ding, Zhenxing Sun
Summary: This paper proposes a visual-aided shared-control method for a semi-autonomous underwater vehicle (sAUV) to conduct flexible, efficient and stable underwater grasping. The method utilizes an arbitration mechanism to assign the authority weights of the human command and the automatic controller according to the attraction field (AF) generated by the target objects and fuses the remote-operation command with a visual servo controller. It has been validated by simulation and experiment to achieve faster and accurate dynamic positioning and reduce the average time consumption on grasping tasks.
JOURNAL OF MARINE SCIENCE AND ENGINEERING
(2023)
Article
Automation & Control Systems
Hoosang Lee, Daehyeon Jeong, Hongje Yu, Jeha Ryu
Summary: Fisheries are crucial for protein supply in the economy. Using autonomous underwater vehicles (AUVs) to detect damaged fishnets offers an efficient and safe solution in turbid underwater environments. This study proposes an AUV pose control strategy based on image mean gradient feature and a convolutional neural network (CNN) combined with a controller, enabling clear net images acquisition and damage inspection in turbid water.
INTERNATIONAL JOURNAL OF CONTROL AUTOMATION AND SYSTEMS
(2022)
Article
Engineering, Marine
Valery Bobkov, Antonina Shupikova, Alexander Inzartsev
Summary: This article focuses on the inspection of underwater pipelines using autonomous underwater vehicles (AUVs), with an emphasis on the efficient use of stereo images for video navigation. The proposed method highlights visible boundaries and calculates the centerline of the pipelines using combined processing of 2D and 3D video data. The study shows that this technology can be practically applied in UP inspection without the need for additional equipment.
JOURNAL OF MARINE SCIENCE AND ENGINEERING
(2023)
Article
Computer Science, Artificial Intelligence
Rui Wang, Liqiang Tang, Yongliang Yang, Shuo Wang, Min Tan, Cheng-Zhong Xu
Summary: In this paper, a control scheme is developed for the trajectory tracking problem of underactuated autonomous underwater vehicles with input saturation, parameter uncertainty, and disturbance. A novel continuous desired heading angle design is proposed, along with dynamic surface control and compensation filter to handle input saturation. Additionally, novel velocity compensation laws are designed to compensate for the sway velocity and handle parameter uncertainty. Simulation results confirm the effectiveness of the proposed method.
IEEE TRANSACTIONS ON INTELLIGENT VEHICLES
(2023)
Article
Engineering, Marine
Sijie Hu, Rendong Feng, Zhanglin Wang, Chengcheng Zhu, Zhikun Wang, Ying Chen, Haocai Huang
Summary: This study successfully addresses the issue of internal inspection in underwater pipeline maintenance. A depth-fixed line tracking control solution is proposed using a circular dish-shaped underwater helicopter (AUH) and an underwater image processing algorithm is utilized to identify the pipeline line accurately.
Article
Engineering, Marine
Tianxing Xia, Dehao Cui, Zhenzhong Chu, Xing Yu
Summary: In order to tackle the challenges of autonomous navigation of unmanned underwater vehicles (UUVs) in long-distance underwater tunnel detection tasks and improve the control performance of their heading control system, a method of autonomous heading planning and control based on sonar-ranging feedback control is proposed. This method combines UUV's autonomous heading planning technology with the heading proportion-integral-derivative (PID) control algorithm, optimizing the acquisition method of controller input data, to provide specific adaptive characteristics to the controller. By utilizing the ranging principle of ultrasonic spontaneous self-collection, the method can obtain the yaw direction and angle of the vehicle relative to the target heading in the tunnel, and continuously adjust the control law to change the heading as the vehicle's heading status changes during navigation. The effectiveness of this method is verified through pool experiments, which show that it achieves good control effect in UUV's underwater tunnel detection heading control and exhibits advantages in long-distance closed tunnel environments. UUV can adaptively adjust the heading according to the tunnel environment and has a fast response and strong applicability in planning and controlling the heading.
JOURNAL OF MARINE SCIENCE AND ENGINEERING
(2023)
Article
Engineering, Civil
Yukiyasu Noguchi, Toshihiro Maki
Summary: This article introduces a method for tracking underwater structures using an AUV, based on stochastic means and utilizing commercial off-the-shelf sensors. It is suitable for surveying various underwater structures and is applicable to fast moving vehicles. The method has been successfully tested in both sea and tank environments, demonstrating its effectiveness in tracking rugged seafloors and vertical walls.
IEEE JOURNAL OF OCEANIC ENGINEERING
(2021)
Article
Engineering, Marine
Tu Lv, Mingjun Zhang, Yujia Wang
Summary: This paper proposes a prediction-based region tracking control scheme for an autonomous underwater vehicle (AUV) to solve the problems of overshoot and high energy consumption. By predicting the future position of AUV and the outer boundary of the desired region, the controller is designed accordingly. Furthermore, an optimization scheme considering the desired region is proposed to address the output saturation issue.
JOURNAL OF MARINE SCIENCE AND ENGINEERING
(2022)
Article
Robotics
Shuo Xu, Yanqing Jiang, Ye Li, Bo Wang, Tianqi Xie, Shuchang Li, Haodong Qi, Ao Li, Jian Cao
Summary: This research proposes a vision-based navigation strategy for AUVs to independently identify and reconstruct the docking station. The proposed framework includes methods for light beacon detection, matching, and fusion pose estimation, which can provide reliable docking navigation for AUVs.
JOURNAL OF FIELD ROBOTICS
(2023)
Article
Chemistry, Analytical
Yongji Zhang, Yu Jiang, Hong Qi, Minghao Zhao, Yuehang Wang, Kai Wang, Fenglin Wei
Summary: The marine environment poses unique challenges for human-robot interaction, especially in underwater gesture recognition. A visual-textual model (VT-UHGR) is proposed in this paper to overcome the difficulties caused by light refraction and color attenuation. By encoding the visual and textual features of underwater divers, the VT-UHGR model generates multimodal interactions to guide AUVs in learning and inference. The results show that incorporating textual patterns significantly improves the performance of underwater gesture recognition.
Article
Engineering, Marine
Ziyi Su, Xiaogong Lin, Bing Huang, Dawei Zhao, Han Sun
Summary: This article investigates an improved adaptive dynamic event-triggered control scheme for autonomous underwater vehicles (AUVs) under the quaternion-based attitude representation. The main innovation is to improve the dynamic event-triggered algorithm to achieve a better trade-off between control performance and communication frequency in the controller-to-actuator channel. A unified auxiliary variable is developed to describe the attitude representation of the AUV, and an event-based adaptive control law is designed to mitigate external perturbations and unknown model dynamics. The trigger frequency can be adapted to the AUV system's performance requirements using a modified trigger adjustment algorithm.
Article
Engineering, Marine
Ioannis Polymenis, Maryam Haroutunian, Rose Norman, David Trodden
Summary: Underwater vehicles have become more sophisticated due to advancements in underwater operations. This study utilizes recent advancements in deep learning to construct a bespoke dataset for underwater applications using generative adversarial networks.
JOURNAL OF MARINE SCIENCE AND ENGINEERING
(2022)
Article
Engineering, Mechanical
Caoyang Yu, Yiming Zhong, Lian Lian, Xianbo Xiang
Summary: This paper proposes a two-layer cascade tracking controller and a deadzone compensator for simplified and effective surge-heading control of underwater vehicles equipped with an X-shaped horizontal actuation configuration. The first-layer cascade system is used for simplified dynamics tracking, while the second-layer cascade system utilizes a reduced-order extended state observer to estimate the uncertainty of the dynamics. Additionally, a dedicated dead-zone compensator is proposed for the X-shaped actuation configuration and the input-to-state stability of the whole tracking system is analyzed.
NONLINEAR DYNAMICS
(2023)
Article
Business
Abdul Qayyum, Imran Razzak, Aamir Saeed Malik, Sajid Anwar
Summary: This study presents a cost-effective framework based on UAV and satellite stereo images to monitor trees and vegetation, providing better disparity. By fusing convolutional neural network and sparse representation, the method can improve vegetation threat estimation accuracy.
TECHNOLOGICAL FORECASTING AND SOCIAL CHANGE
(2021)
Article
Computer Science, Theory & Methods
Abdul Qayyum, Moona Mazher, Aliyu Nuhu, Abdesslam Benzinou, Aamir Saeed Malik, Imran Razzak
Summary: The paper presents a health monitoring method using remote photoplethysmography (rPPG) and a smartphone camera to estimate vital signs. By reconstructing a sparse signal, the method accurately estimates physical parameters such as heart rate, breathing rate, heart rate variability, and SpO2. The power ratio in the frequency domain of inter-beat intervals is used to measure physiological states like stress and fatigue.
Article
Chemistry, Analytical
Zahid Mahmood, Khurram Khan, Uzair Khan, Syed Hasan Adil, Syed Saad Azhar Ali, Mohsin Shahzad
Summary: This paper proposes an efficient license plate detection method by combining Faster R-CNN with digital image processing techniques. The method first detects vehicles using Faster R-CNN and then analyzes the located vehicle using a robust License Plate Localization Module. The module uses color segmentation, HSV image processing, morphological filtering, and dimension analysis to detect the license plate and achieves high accuracy in less execution time.
Article
Computer Science, Artificial Intelligence
Ubaid M. Al-Saggaf, Syed Faraz Naqvi, Muhammad Moinuddin, Sulhi Ali Alfakeh, Syed Saad Azhar Ali
Summary: Mental stress has severe effects on physical and psychological health, making timely diagnosis and assessment crucial. Currently, there is no wearable or portable device developed specifically for stress assessment, which requires a time-efficient algorithm. This study compared machine learning and deep learning approaches in terms of time required for feature extraction and classification, finding that deep learning provides automated unsupervised feature extraction and efficient classification, making it suitable for real-time mental stress assessment in wearable devices.
FRONTIERS IN NEUROROBOTICS
(2022)
Review
Environmental Sciences
Rumaisa Abu Hasan, Muhamad Saiful Bahri Yusoff, Tong Boon Tang, Yasir Hafeez, Mazlina Che Mustafa, Masayu Dzainudin, Juppri Bacotang, Ubaid M. Al-Saggaf, Syed Saad Azhar Ali
Summary: This review examines the overall effectiveness of resilience-building interventions conducted on teachers in early childhood education (ECE) settings. The findings suggest a preference for group approaches and varying durations of interventions. Challenges with the group approach, including lengthy interventions and participant attrition, are highlighted. Additionally, there is a need for physiological measures to evaluate the effects on mental health.
INTERNATIONAL JOURNAL OF ENVIRONMENTAL RESEARCH AND PUBLIC HEALTH
(2022)
Article
Green & Sustainable Science & Technology
Mohsin Shahzad, Arsalan Qadir, Noman Ullah, Zahid Mahmood, Naufal Mohamad Saad, Syed Saad Azhar Ali
Summary: This paper presents an optimized solution for the existing power system in the Azad Jammu and Kashmir region, utilizing a renewable energy-based generation system to address the challenges faced by the power system. Through comparisons with other models, the proposed hybrid energy generation system demonstrates its techno-economic benefits by reducing energy costs, net present costs, initial costs, and greenhouse gas emissions, while improving the voltage profile of the system.
Article
Chemistry, Multidisciplinary
Salman A. Khan, Kashif Iqbal, Nazeeruddin Mohammad, Rehan Akbar, Syed Saad Azhar Ali, Ammar Ahmed Siddiqui
Summary: This paper proposes a new evaluation metric for email spam detection based on fuzzy logic concept, and it confirms the effectiveness through empirical analysis and extrinsic evaluation.
APPLIED SCIENCES-BASEL
(2022)
Article
Chemistry, Analytical
Atif Anwer, Samia Ainouz, Mohamad Naufal Mohamad Saad, Syed Saad Azhar Ali, Fabrice Meriaudeau
Summary: Detecting and removing specular highlights in images is a significant problem. Existing techniques are inadequate for real-world images with complex textures and multiple objects. This paper proposes an efficient Specular Segmentation network based on the U-net architecture, which can accurately detect specular pixels in various real-world images.
Article
Chemistry, Multidisciplinary
Muhammad Waheed Sabir, Zia Khan, Naufal M. Saad, Danish M. Khan, Mahmoud Ahmad Al-Khasawneh, Kiran Perveen, Abdul Qayyum, Syed Saad Azhar Ali
Summary: Image segmentation is a common task in medical image analysis, and the segmentation of the liver and liver tumors is crucial for screening and diagnosing liver diseases. In this article, the ResU-Net network is implemented to segment the liver and tumors from CT scan images, with pre-processing techniques applied for image quality enhancement. The network performance is evaluated using the Dice similarity coefficient.
APPLIED SCIENCES-BASEL
(2022)
Article
Pathology
Amjad Khan, Nelleke Brouwer, Annika Blank, Felix Mueller, Davide Soldini, Aurelia Noske, Elisabeth Gaus, Simone Brandt, Iris Nagtegaal, Heather Dawson, Jean-Philippe Thiran, Aurel Perren, Alessandro Lugli, Inti Zlobec
Summary: This study proposes a deep learning-based workflow for evaluating lymph node metastases in colorectal cancer. The method achieved high accuracy after training on a dataset of 100 whole-slide images. The fine-tuned models showed significant improvements in metastasis detection and achieved excellent performance in the validation cohorts.
Article
Multidisciplinary Sciences
Usman Qamar Shaikh, Muhammad Shahzaib, Sadia Shakil, Farrukh A. Bhatti, Aamir Saeed Malik
Summary: This study presents an inertial sensor-based approach for real-time gait event detection system that works for diverse terrains. The system has high accuracy and computational efficiency, making it suitable for use in dynamic environments.
Proceedings Paper
Computer Science, Artificial Intelligence
Vojtech Mrazek, Soyiba Jawed, Muhammad Arif, Aamir Saeed Malik
Summary: In this paper, an interpretable EEG-based solution for the diagnostics of MDD is proposed, which involves the acquisition of EEG data from MDD patients and healthy controls. The best features are selected using the NSGA-II algorithm and utilized for SVM and k-NN classifiers. The results show that gamma bands extracted from the left temporal brain regions can significantly distinguish MDD patients from controls. The proposed solution achieves high sensitivity, specificity, and accuracy.
PROCEEDINGS OF THE 2023 GENETIC AND EVOLUTIONARY COMPUTATION CONFERENCE, GECCO 2023
(2023)
Proceedings Paper
Computer Science, Information Systems
H-G Nguyen, A. Khan, H. Dawson, A. Lugli, I Zlobec
Summary: This study proposes a group affinity weakly supervised segmentation method (GAWS) for precise tissue segmentation of histopathology images. It creates cluster and target images by extracting visual features and applying constraints, and updates network parameters through backpropagation. The method shows excellent inter-observer agreement and accuracy in quantification of extracellular mucin-to-tumor area.
MEDICAL IMAGING 2022: DIGITAL AND COMPUTATIONAL PATHOLOGY
(2022)
Proceedings Paper
Computer Science, Information Systems
Mauro Gwerder, Amjad Khan, Christina Neppl, Inti Zlobec
Summary: This paper explores a weakly supervised learning algorithm for computational tools in pathology. By training on whole slide images of lymph nodes from lung cancer patients, the algorithm achieves high accuracy and AUC. Optimizing the model through smart selection of relevant tiles reduces the amount of training data required.
MEDICAL IMAGING 2022: DIGITAL AND COMPUTATIONAL PATHOLOGY
(2022)
Proceedings Paper
Engineering, Biomedical
Amjad Khan, Manfredo Atzori, Sebastian Otalora, Vincent Andrearczyk, Henning Mueller
Summary: The study compares methods for dealing with stain color heterogeneity in histopathology slides to improve machine learning-based computational analysis in normal versus tumor tissue classification. Through systematic experimentation, stain color normalization and augmentation techniques are used to train CNNs to generalize on unseen data from multiple centers, resulting in improved performance metrics on external test sets.
MEDICAL IMAGING 2020: DIGITAL PATHOLOGY
(2021)