A Robust Real-Time Detecting and Tracking Framework for Multiple Kinds of Unmarked Object
Xiaodong Lv1, Chuankai Dai1, Luyao Chen2
1Beijing Advanced Innovation Center for Intelligent Robot and System, Beijing Institute of Technology, Beijing 100081, China.
Sensors (Basel, Switzerland)
|December 22, 2019
Summary
This study introduces a novel rodent real-time tracking framework using deep learning (You Only Look Once, v3) and Kalman Filter. The combined system significantly improves tracking accuracy for multiple rodents compared to standalone methods.
Area of Science:
- Computer Vision
- Animal Behavior Analysis
- Machine Learning
Background:
- Accurate real-time tracking of multiple rodents is crucial for behavioral studies.
- Existing methods often struggle with occlusions and maintaining object identity.
Purpose of the Study:
- To develop an automated framework for real-time detection and tracking of multiple rodents.
- To enhance tracking accuracy and robustness by integrating deep learning with predictive algorithms.
Main Methods:
- A deep learning model (Rat-YOLO v3) was trained for object detection.
- Kalman Filter was employed for position prediction and ID association.
- An improved Hungarian algorithm was used for object matching between frames.
- A nine-point position correction algorithm refined detection accuracy.
Main Results:
- The standalone Rat-YOLO model achieved 72.545% accuracy in detecting three rats.
- The integrated framework (Rat-YOLO with Kalman Filter and nine-point correction) improved accuracy to 95.194%.
Conclusions:
- The proposed integrated framework significantly enhances the accuracy and reliability of multi-rodent real-time tracking.
- This system offers a robust solution for automated behavioral analysis in research settings.
