Motion Image-based Mouse Tracking Algorithm for Morris Water Maze Test

摘要

We propose two techniques for tracking a mouse in a Morris water maze test by using motion image processing: first, a calculating technique of the path-length for tracking a mouse in the maze, and secondly, a confirmation technique of a mouse's arrival at the platform. The former can track the mouse without special conditions for avoiding artifacts, such as those from reflected lights. The latter can achieve remarkable accuracy by using the area-ratio of a mouse shape instead of center-of-gravity. The proposed techniques are very useful for analyzing the track of a mouse in a Morris water maze test.