Sensor Fusion-Based Vehicle Detection and Tracking Using a Single Camera and Radar at a Traffic Intersection.

Area of Science:

• Intelligent Transportation Systems (ITS)
• Computer Vision
• Sensor Fusion

Background:

• Real-time traffic control systems benefit from advanced sensor technologies, signal processing, and machine learning.
• Adapting to dynamic traffic conditions requires efficient vehicle detection and tracking.

Purpose of the Study:

• To introduce a novel, cost-effective sensor fusion approach for vehicle detection and tracking.
• To integrate data from a single camera and radar for enhanced traffic monitoring.

Main Methods:

• Independent vehicle detection and classification using camera and radar data.
• Kalman filter with a constant-velocity model for predicting vehicle locations.
• Hungarian algorithm for associating predictions with sensor measurements.
• Merging kinematic information for robust vehicle tracking.

Main Results:

• Demonstrated effectiveness of the sensor fusion method in a real-world intersection case study.
• Achieved efficient and accurate vehicle detection and tracking.
• Outperformed individual sensor performance in comparative analysis.

Conclusions:

• The proposed camera-radar sensor fusion method provides a cost-effective and efficient solution for real-time traffic monitoring.
• This approach significantly improves vehicle detection and tracking capabilities for intelligent transportation systems.