A5059079323- Underwater Vehicles and Communication Systems
- Water Quality Monitoring Technologies
- Underwater Acoustics Research
- Robotics and Sensor-Based Localization
- Robotic Path Planning Algorithms
- Maritime and Coastal Archaeology
- Maritime Navigation and Safety
- Smart Agriculture and AI
- Modular Robots and Swarm Intelligence
- Adaptive Control of Nonlinear Systems
- Advanced Control Systems Optimization
- Marine Ecology and Invasive Species
- Infrared Target Detection Methodologies
- Neural Networks and Applications
- Evolutionary Algorithms and Applications
- Fuzzy Logic and Control Systems
- Control and Dynamics of Mobile Robots
- Microplastics and Plastic Pollution
- Insect Pheromone Research and Control
- Marine Bivalve and Aquaculture Studies
- Conservation Techniques and Studies
- Identification and Quantification in Food
- 3D Surveying and Cultural Heritage
- Structural Integrity and Reliability Analysis
- Infrastructure Maintenance and Monitoring
University of Zagreb
2017-2024
University of Sarajevo
2014-2015
Aquaculture net pens inspection and monitoring are important to ensure stability fish health in the farms. Remotely operated vehicles (ROVs) offer a low-cost sophisticated solution for regular of underwater due their ability visual sensing autonomy challenging dynamic aquaculture environment. In this paper, we report integration an ROV with servoing scheme tracking pens. We propose vision-based positioning that consists object detector, pose generator, closed-loop controller. The system...
There are activities in viticulture and mariculture that require extreme physical endurance from human workers, making them prime candidates for automation robotization. This paper presents a novel, practical, heterogeneous, autonomous robotic system divided into two main parts, each dealing with respective scenarios mariculture. The components the subsystems enable collaboration were developed as part of ongoing HEKTOR project, specific scenario is presented. In viticulture, this includes...
The 2 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">nd</sup> Workshop on Maritime Computer Vision (MaCVi) 2024 addresses maritime computer vision for Unmanned Aerial Vehicles (UAV) and Surface (USV). Three challenges categories are considered: (i) UAV-based Object Tracking with Re-ideruification, (ii) USV-based Obstacle Segmentation Detection, (iii) Boat Tracking. Detection features three sub-challenges, including a new embedded challenge...
Underwater cultural heritage sites are subject to constant change, whether due natural forces such as sediments, waves, currents or human intervention. Until a few decades ago, the documentation and research of these was mostly done manually by diving archaeologists. This paper presents results integration remote sensing technologies with autonomous marine vehicles in order make task site even faster, more accurate, efficient precisely georeferenced. It includes multibeam sonar, side scan...
In the interest of both enabling long-term autonomous monitoring at-risk marine environments and raising awareness capabilities among citizens, a heterogeneous system robots was developed, integrated, deployed on mission in Adriatic Sea. This paper details use-case scenario for team robotic agents purpose cooperative litter detection mapping, while also including interested citizens loop allowing them to serve as operators. Two Autonomous Surface Vehicles (ASVs), Remotely Operated Vehicle...
This paper presents the overview and preliminary results of HEKTOR - Heterogeneous Autonomous Robotic System in Viticulture Mariculture project. is divided into two main parts, each dealing with specific scenarios viticulture mariculture. The robots used project scenario considered are presented. In viticulture, this includes vineyard surveillance, spraying bud rubbing using an all-terrain mobile manipulator unmanned aerial vehicle (UAV). mariculture, include coordinated monitoring fish net...
The process of fish cage inspections, which is a necessary maintenance task at any farm, be it small-scale or industrial, that has the potential to fully automated. Replacing trained divers who perform regular inspections with autonomous marine vehicles would lower costs manpower and remove risks associated humans performing underwater inspections. Achieving such level autonomy implies developing an image processing algorithm capable estimating state biofouling buildup. aim this work propose...
A linear model predictive control (LMPC) based framework is developed for underactuated marine vehicles' kinematic line following while moving at a constant depth in the presence of disturbance. LMPC used as high level controller which sets reference surge and yaw velocities low PID tracking controllers. Used ACADO optimization toolbox solves optimal problem hand real-time. This has shown good results both simulation environment on-sea experiments.
Plitvice Lakes National Park is the largest national park in Croatia and also oldest from 1949. It was added to UNESCO World Natural Heritage List 1979, due unique physicochemical biological conditions that have led creation of 16 named several smaller unnamed lakes, which are cascading one into next. Previous scientific research proved increased amount dissolved organic matter (pollution) stops travertine processes on Lakes. Therefore, this complex, dynamic but fragile geological,...
This paper presents an overview and preliminary results of the HEKTOR (Heterogeneous Autonomous Robotic System in Viticulture Mariculture) project. An survey applications a heterogeneous cooperative autonomous robotic system, consisting aerial, surface underwater vehicles, mariculture scenarios is presented. Target system are fish net cage inspection (biofouling damage detection) as well biomass estimation. Furthermore, detailed description acquired vehicles given, namely unmanned aerial...
An autonomous underwater vehicle (AUV) or a multi-AUV system performing seafloor exploration missions by side-scan sonar need to perceive their environment in order replan the mission if they detect interesting objects sensor data. Several anomalous/salient object detection methods mostly used for natural images are here applied images. All were firstly benchmarked on 1500 simulated dataset. Precision-recall and processing time analysis was conducted choose best suited method such controlled...
ooperation of a remotely operated vehicle (ROV) and an autonomous surface (ASV) in prolonged robotic inspections missions mariculture requires tether management system (TMS) with docking mounted onto the ASV. The developed scope HEKTOR (Heterogeneous Autonomous Robotic System Viticulture Mariculture) project is shortly presented. Furthermore, mechanical electrical design TMS, control scheme communication protocol TMS ASV, TMS-ROS interface for future high-level mission are described....
Cooperation of a lightweight autonomous aerial vehicle (LAAR) and an surface (ASV) in prolonged robotic inspections missions mariculture requires landing platform (LP) with docking system mounted onto the ASV. The vehicles developed scope HEKTOR (Heterogeneous Autonomous Robotic System Viticulture Mariculture) project are presented herein. Furthermore, mechanical electrical design LP, control scheme communication protocol LP ASV, LP-ROS interface for future high-level mission described. ASV...
This paper deals with the development of control software for a remotely operated vehicle (ROV) as part an automated fish cage inspection system in aquaculture. The ROV navigates autonomously around cages streaming video to topside computer that runs algorithms vertical rope detection. sends back velocity references real-time order successfully complete task. These images are also used determine state net biofouling using pre-trained convolutional neural network help which nets need...
A modular measurement model Extended Kalman filter (EKF) for unmanned underwater vehicle (UUV) localization is proposed. Except using measurements from UUV's sensors, this EKF augmented by ultra-short baseline range and visual-data based an surface vehicle, in-sonar image estimated UUV position. It shown that the proposed significantly enhances navigational accuracy through a collaborative fusion of sensor data multiple heterogeneous marine vehicles. Also, Rauch-Tung-Striebel (ERTS) smoother...
Side-scan sonar mapping of an unknown large-scale seafloor area by a marine vehicle is nowadays very common. It also important that a-priori interesting parts the are scanned in more detail, i.e. sonified from both sides. However, completely autonomous and time-efficient coverage path (re)planning for such missions still open issue. In contrast to standard overlap-all-sonar-ranges lawnmower pattern offline static problem solution side-scan missions, this paper two online data-driven...
This paper presents an overview of advances in estimation the biofouling state fish cages as a part HEKTOR (Heterogeneous Autonomous Robotic System Viticulture and Mariculture) project. Firstly, developed framework for is shown explained brief. A method using k-means clustering labeling images outlined. It followed by results machine learning approaches automatic inferring semantic meaning pixels image trained on recorded dataset said outlined method. Furthermore, brief contour detection...
The process of fish cage inspections, which is a necessary maintenance task at any farm, be it small scale or industrial, that has the potential to fully automated. Replacing trained divers who perform regular inspections with autonomous marine vehicles would lower costs manpower and remove risks associated humans performing underwater inspections. Achieving such level autonomy implies developing an image processing algorithm capable estimating state biofouling buildup. aim this work propose...
It is well known that the process of tuning a fuzzy logic controller almost always very complex task, which time consuming, laborious and often requires expert knowledge controlled system. Mapping controller's parameters (rule base membership functions input(s) output(s)) into performance measure in closed analytical form near impossible to get, thus use any classical optimization method automatically ruled out. Knowing this, genetic algorithms with fitness function cumulative response error...
The 2nd Workshop on Maritime Computer Vision (MaCVi) 2024 addresses maritime computer vision for Unmanned Aerial Vehicles (UAV) and Surface (USV). Three challenges categories are considered: (i) UAV-based Object Tracking with Re-identification, (ii) USV-based Obstacle Segmentation Detection, (iii) Boat Tracking. Detection features three sub-challenges, including a new embedded challenge addressing efficicent inference real-world devices. This report offers comprehensive overview of the...