Related Experiment Video
Updated: Oct 13, 2025
08:32
Tracking Rats in Operant Conditioning Chambers Using a Versatile Homemade Video Camera and DeepLabCut
Published on: June 15, 2020
12.8K
A robust bitmap-based real-time position tracking algorithm for rats in radial arm maze tests
Chi-Chun Chen1, Liang-Jyun Hong2, Jian-Yong Wang2
1Department of Electronic Engineering, National Chin-Yi University of Technology, Taichung, Taiwan. chichun@ncut.edu.tw.
Scientific Reports
|November 18, 2021
Summary
This study presents a fast and accurate real-time rat position tracking algorithm for radial arm mazes. The developed method significantly outperforms YOLO v2 in speed while maintaining high accuracy, even with environmental disturbances.
Area of Science:
- Animal behavior research
- Computer vision applications
- Robotics and automation
Background:
- Accurate real-time tracking of animal subjects in behavioral experiments is crucial for data integrity.
- Existing methods may lack robustness or computational efficiency, limiting their practical application.
- Radial arm mazes are common tools for studying spatial learning and memory in rodents.
Purpose of the Study:
- To develop a novel, highly accurate, and computationally efficient real-time position tracking algorithm for rats in a radial arm maze.
- To evaluate the algorithm's robustness against environmental noise and external interference.
- To compare the performance of the developed algorithm against a state-of-the-art deep learning model (YOLO v2).
Main Methods:
- Utilized an infrared night-vision camera for image acquisition above the radial arm maze.
- Applied image processing techniques including binarization, intersection, and opening operations.
- Tested algorithm performance with and without an experimenter intruding the maze.
- Compared processing time and accuracy against a pre-trained YOLO v2 model.
Main Results:
- Achieved over 99% accuracy in real-time rat position tracking under normal conditions.
- Demonstrated high robustness, with accuracy only slightly decreasing to 98% in the presence of an intruder.
- The developed algorithm processed over 9000 frames in minutes, significantly faster than YOLO v2's hour-long processing time.
- Outperformed YOLO v2 in accuracy (99%+ vs 97%+) and demonstrated superior speed.
Conclusions:
- The developed position tracking algorithm offers a robust, accurate, and highly efficient solution for rodent tracking in behavioral studies.
- Its computational efficiency makes it a valuable tool for real-time analysis, overcoming limitations of existing deep learning approaches.
- The algorithm's performance highlights its potential for advancing research in neuroscience and animal behavior.
