Automated tracking of aquatic crustaceans with potential application on the quantification of animals movement

Abstract

Here, we present a set of algorithms using the Python programming language, that will allow using a routine for object detection and tracking in experimental videos. We developed a script, under the fundamentals of background subtraction and image thresholding (using the OpenCV package), that makes it possible to track a wide spectrum of animals under different conditions. We have validated this script through testing on semi-terrestrial and aquatic crustacean species and under different experimental scenarios (laboratory and field sampling and using video created under nocturnal and diurnal conditions). The open-source nature of the script allows for flexibility and scalability, so it can be easily customized and is thus transferable to other species/experiments in the context of behavioral ecology. The tracking script is easy customizable and free alternative to commercial video tracking systems and therefore, applicable to a wide variety of both educational and research programs.

Keywords

• Python
• OpenCV
• object detection
• video tracking
• behavioural ecology