Automatic animal detection | |
Observation of animal behavior is important to many fields of research in the life sciences. Its automation is crucial to advances in animal well-being and scientific research. In the EthoVision product made by Noldus, a Windows computer is attached to a camera which has a 2D top view of a confined space: The initial setup involves telling the system how to extract the animal from the background. The user defines settings which apply to the whole of the recording period. Typically, he tries several alternative methods for background extraction, each containing various thresholds with which to experiment. It becomes increasingly more difficult to find the optimal settings. | |
Assignment The purpose of this assignment is to eliminate this initial setup. It will be replaced by an automated process based on for instance machine learning, artificial intelligence, or a library based approach. After a training period, during which the system observes the live or recorded environment, the system runs unattended. Optionally, it continues to adjust the settings over time (for instance: circadian) | |
Your background | |
More information If you are interested in this particular project, please contact Rob Ottenhoff (r.ottenhof@noldus.nl)or Wil van Dommelen (w.van.dommelen@noldus.nl). | |

for instance a cage, pool, maze or Petri-dish with animal(s) or insects in it. The images are recorded and analyzed in real-time to generate statistics on what the animals being observed are doing. After an initial set-up, this system typically runs unattended for minutes, hours or days. 