Research on Target Detection Based on Distributed Track Fusion for Intelligent Vehicles

Area of Science:

• Intelligent Transportation Systems
• Sensor Fusion
• Computer Vision
• Radar Technology

Background:

• Accurate target detection is crucial for intelligent vehicle safety.
• Existing sensors have perception-level defects that limit performance.
• Sensor fusion technology offers a solution to compensate for sensor limitations.

Purpose of the Study:

• To investigate the application of sensor fusion in intelligent vehicle target detection.
• To develop and evaluate a novel fusion algorithm combining millimeter-wave radar and camera data.
• To enhance the accuracy and reliability of target detection systems.

Main Methods:

• A target-level fusion hierarchy with a distributed structure was adopted.
• The algorithm includes two tracking processing modules and one fusion center module.
• A multi-target tracking algorithm, regional collision association, weighted track association, and a non-reset federated filter were employed.

Main Results:

• The proposed algorithm successfully achieved track association between millimeter-wave radar and camera data.
• The fusion track state estimation method demonstrated excellent performance.
• The system effectively compensated for individual sensor defects through fusion.

Conclusions:

• The developed sensor fusion algorithm enhances target detection capabilities for intelligent vehicles.
• The combination of millimeter-wave radar and camera data significantly improves system reliability.
• This approach provides a robust solution for safe and autonomous driving.