Research Focus and Directions
The main goal of the laboratory is applied research focused on introducing automation and digitalization into aquaculture and water protection processes. We aim to develop low-cost systems for automatic and non-invasive monitoring that can be applied in practice.
Key activities:
- Development of fish imaging systems
- Remote monitoring of water quality
- Monitoring of aquatic animal behavior (welfare)
- Analysis of fish coloration
- Detection of fish diseases based on image analysis
- Individual identification of fish based on image analysis
- Toxicity and biocompatibility testing (assessment of water cytotoxicity)
- Toxicity and biocompatibility testing (providing information)
We offer free collaboration through open access to our services and research infrastructure.
Key projects:
Individual identification of fish from images
The aim of the project is to replace invasive fish tagging with non-invasive individual identification based on their appearance.
Under laboratory conditions, we have already verified that patterns on fish skin can be used to recognize individual salmon, carp, sea bass, rainbow trout, and tiger barbs. The next step is to implement the method so that it can be used under real conditions in tanks and cages.
Support for ethological studies
Fish behavior reflects their overall condition. If we are able to track changes in behavior or differences between groups, we can infer changes in health or physiological state from these data. A series of projects focuses on the automated parameterization of fish behavior affected by nanoparticles, pharmaceuticals, or changes in welfare conditions.
Image-based analysis of fish diseases
The aim of the project is to create a camera system that can be placed directly in a fish tank and is capable of detecting visual symptoms of diseases on the fish body.
Monitoring of fish passes
Fish passes are structures installed at barriers in rivers that facilitate the natural migration of fish. We have developed a technology for monitoring passing fish in real time using infrared light and camera systems. The research focuses on questions such as how fish passes function, which fish species pass through them, and in what numbers. To answer these questions, we use neural networks for detection and classification.
Estimation of water parameters from satellite images
The quality of inland waters (ponds and reservoirs) is crucial for fish farming as well as for recreational use of water bodies by the public. Standard water quality analysis is based on sampling and laboratory analyses, which are time-consuming and costly.
We have developed a system for estimating basic parameters (e.g., chlorophyll A) from freely available satellite images.
Regenerative medicine
This project focuses on the development of scaffolds from decellularized human tissue for use in regenerative medicine. Key steps include the design and implementation of a decellularization unit using supercritical CO₂, optimization of protocols, and seeding the scaffolds with mesenchymal stem cells.
Automatic detection of living cells
We develop methods for automated cell analysis from microcinematographic recordings using convolutional neural networks (CNN). The optimized CNN model enables detailed analysis of cell interactions with samples, which is crucial for further phases of research.
Cytotoxicity testing of liquid samples
Using advanced image analysis of living cells, we test the cytotoxicity of liquid samples, including leachates and extracts from solid materials. We use time-lapse microcinematography, which enables long-term monitoring of cellular responses without staining, revealing even subtle morphological and growth changes.
Biocompatibility testing of solid samples
We test the biocompatibility of solid samples by monitoring the interactions of living cells with material surfaces. Using time-lapse microcinematography, we analyze cell growth, adhesion, and behavior in contact with the tested material, which allows evaluation of its suitability for biomedical applications.
