Related Experiment Video
Updated: Dec 21, 2025
11:53
The Modular Design and Production of an Intelligent Robot Based on a Closed-Loop Control Strategy
Published on: October 14, 2017
12.1K
Deep-Learning-Based Indoor Human Following of Mobile Robot Using Color Feature
Redhwan Algabri1, Mun-Taek Choi1
1School of Mechanical Engineering, Sungkyunkwan University, Suwon 16419, Korea.
Sensors (Basel, Switzerland)
|May 14, 2020
Summary
This study presents a novel robot framework for reliable human following, even with visual challenges like occlusion and changing light. The system effectively tracks targets and navigates safely to destinations, demonstrating practical application.
Area of Science:
- Robotics
- Artificial Intelligence
- Computer Vision
Background:
- Human following is crucial for mobile robot interaction.
- Challenges include occlusion, illumination changes, and obstacle avoidance.
- Existing methods often struggle with dynamic environments.
Purpose of the Study:
- To develop a robust human following framework for mobile robots.
- To enable target tracking through occlusions and illumination variations.
- To ensure safe navigation with obstacle avoidance during human following.
Main Methods:
- Utilized a state-machine control framework.
- Employed deep learning (Single Shot MultiBox Detector) for person detection and tracking.
- Used hue-saturation-value histogram for target identification.
- Implemented simultaneous localization and mapping (SLAM) with LIDAR for navigation and obstacle avoidance.
Main Results:
- The robot successfully tracked the target person in an indoor environment.
- The system demonstrated effective performance despite moderate illumination changes and multiple people.
- The robot navigated safely to the destination while maintaining target following.
Conclusions:
- The proposed framework is effective and practical for mobile robot human following.
- The integration of deep learning and SLAM enables robust performance in challenging conditions.
- This system enhances human-robot interaction capabilities for mobile robots.
