1 Noldus Information Technology b.v., Wageningen, The Netherlands
2 Denka International b.v., Barneveld, The Netherlands
3 ID-DLO, Lelystad, The Netherlands
4 Department of Mathematics, University of Leiden, Leiden, The Netherlands
During the past years, video tracking has become a standard method in neurological, pharmacological and entomological research. This technique uses characteristics of locomotion patterns to quantify behavior. Due to technical and practical limitations, the application of this technique has been restricted to observation of a single unmarked animal or a small number of marked animals (up to 8). Techniques such as partially coloring animals black and applying colored markers to the animal are used to archieve identification. Without the use of these marking techniques, animals cannot be identified individually and reconstruction of the paths is not possible.
In many entomological applications, a large number of similar-looking insects are used at the same time. In such studies, the researcher is only interested in general characteristics of the observation (number of insects that visited a certain area, average movement speed in a certain area) and not in a detailed study of individual insects. An example is the study of the efficiency of sticky traps. In this case the researcher uses the latency to the trap as a measure for the effectivity of the bait and counts the number of escapes from the trap as a measure for the quality for the used glue.
Traditional tracking systems are unsuitable for this kind of locomotory studies since marking of insects is labour-intensive and difficult, or even impossible due to small size. As a solution, a video tracking system has been designed that is capable of tracking large amounts of unmarked insects simultaneously. In this system a novel approach is used to recombine individual insect positions into a series of paths. The method used for path reconstruction is based on determining the mathematical cost for each combination of a newly measured postition and the already reconstructed paths. This cost is based on differences in position, velocity and walking direction. By minimizing the cost between subsequent measurements, the paths are reconstructed.
This poster presents an overview of the techniques that are used to reconstruct the paths and discusses the advantages and disadvantages. Also the first measurements performed with the system in an experimental setup are presented.