Three-Dimensional Motion Tracking System for Extracting Spatial Movement Pattern of Small Fishes

The mathematical properties of the spatial movement patterns of animals, humans, and insects have gradually become clear in recent years. Motion tracking is essentially necessary for the study of the spatial movement patterns. GPS telemetry is often used for large mammals, birds, and humans. However, it is difficult to track the migration paths of insects and small fish by using GPS telemetry. When the region of an object’s movement is restricted, we can record the movement of the object by the video instead of GPS telemetry, but we must determine the position coordinates of objects from a video for the study of the spatial movement patterns. If this motion tracking can be executed not by manually but automatically, we can obtain and analyze Big data on motion. In this paper, we develop a system for motion tracking of one or more small fishes in an aquarium. This system solves the difficulties such as the overlap of fishes, the ghost image of the reflection, and outputs the trajectory of the 3-dimensional coordinates. Furthermore, we apply this system to actual videos and show that the detection of active and inactive phases is possible and that the spatial movement pattern follows Levy walk.