|
|
| Name | Tomasz Pinkiewicz |
|---|---|
| Position | PhD |
| Qualifications | BCompHons(1) |
| tomaszp@utas.edu.au | |
| Campus | Launceston |
| Phone | (03) 6324 3902 |
| Room | V178 |
| About Me | Thesis: Computational Techniques for Automated Tracking and Analysis of Fish Movement in a Controlled Aquatic Environment Supervisors: Dr. Ray Williams, School of CIS, Dr. John Purser, NCMCRS, AMC Research Group: Marine Technology Hardware/Software: MATLAB, Simulink, C/C++, image and video processing, underwater video camera Description: In a commercial aquaculture farm, knowledge about fish behaviour can influence farm management, fish growth and operating costs. Observing fish movement and identifying behaviours can be a very tedious task both in commercial and research environments. Currently it's only done for short periods of time an on an ad hoc basis using video based technology. My project attempts to automate this process both in sea cages and in small research tanks. This will enable farm operators and researchers to carry out long term analysis of fish movements and provide data to detect daily and seasonal behavioural patterns. Because the project leverages already widely used video technology, it intends to add value to the current farm management partices and to video based aquaculture experiments. Analysis in sea cages involves large numbers of fish, so the aim is to only track fish for short periods of time. Sampling every 30 seconds, I calculate an average speed and direction of movement and this allows me to create a daily swimming profile. Because swimming velocity is influenced by environmental conditions, daily water temperatures and dissolved oxygen levels are recorded as well as times of high/low tides. In addition feeding times are recorded to allow analysis of fish movement during feeding as opposed to non-feeding. In research tanks only a small number of fish are involved. The aim is to track individuals for long periods of time (8-12 hours) in order to determine interactions between individuals. It is also important to determine spatio-temporal distributions of fish without tracking individuals and analyse these distributions in relation to feeding, time of day and other experimental settings. Observing fish in small tanks allows researchers to learn about how fish interact, and determine their learning abilities. |