Related Experiment Video
Updated: Jul 3, 2025
08:32
Tracking Rats in Operant Conditioning Chambers Using a Versatile Homemade Video Camera and DeepLabCut
Published on: June 15, 2020
12.5K
Research on Multi-Hole Localization Tracking Based on a Combination of Machine Vision and Deep Learning
Rong Hou1, Jianping Yin1, Yanchen Liu1
1School of Mechanical and Electrical Engineering, North University of China, Taiyuan 030051, China.
Sensors (Basel, Switzerland)
|February 10, 2024
Summary
This study introduces a vision-based system using YOLOv5 for precise robotic arm control in industrial settings. The method enhances efficiency and safety by automating workpiece assembly and reducing human risk.
Area of Science:
- Robotics
- Computer Vision
- Artificial Intelligence
Background:
- Manual industrial assembly is inefficient and poses risks.
- Traditional machine learning struggles with complex industrial environments.
- Environmental changes impact robot accuracy.
Purpose of the Study:
- To develop an automated system for workpiece localization and assembly.
- To improve industrial robot anti-interference capabilities and efficiency.
- To reduce human exposure to hazardous assembly tasks.
Main Methods:
- Utilizing machine vision and the YOLOv5 deep learning model for target recognition.
- Employing ROS (Robot Operating System) for robotic arm control via coordinate mapping.
- Implementing hand-eye calibration for accurate coordinate transformation.
Main Results:
- High accuracy achieved in target object training and testing.
- Demonstrated high control accuracy for the robotic arm.
- The system exhibits strong anti-interference capabilities in complex industrial environments.
Conclusions:
- The proposed method is feasible and effective for automated industrial assembly.
- It enhances environmental adaptability and work efficiency.
- It provides a viable solution for automated docking disk workpiece installation and screw positioning.
