Thesis supervisor: Máté Nagy
Location of studies (in Hungarian): ELTE Department of Biological Physics Abbreviation of location of studies: ELTE
Description of the research topic:
The conventional technique to estimate population sizes of bird species at roost sites is done by an observer who manually count the birds as they enter or leave. The rapidly improving computer vision techniques - machine learning, deep neural networks - could be of help to quantify the number of birds on a given video frame, but estimation of the full population is still not trivial due to the perhaps millions of birds gathered in a dense and relatively localised area. What potential camera setup would be the most suitable for this task? Could it be a fleet of drones equipped with cameras or perhaps an array of traditional cameras mounted at fixed locations? By simulating bird flocks and different camera setups, it will be possible to test the viability of different image-capturing techniques before actually carrying out experimental data collection. Knowing the population size is an important tool for planning conservation actions. Given the interest of the candidate, the results of the modelling could be used first hand in an experimental study on a large hub of red-footed falcons in Angola.
Required language skills: English Further requirements: An ideal candidate for this project has good quantitative skills for modelling and data analysis, and strong commitment to work on biologically interesting problems. This collaborative work will include Dr Emil Andersson and Dr Péter Palatitz (BirdLi
Number of students who can be accepted: 2
Deadline for application: 2024-05-31
2024. IV. 17. ODT ülés Az ODT következő ülésére 2024. június 14-én, pénteken 10.00 órakor kerül sor a Semmelweis Egyetem Szenátusi termében (Bp. Üllői út 26. I. emelet).
Login
