Effective vehicle-to-vehicle positioning method using monocular camera based on VLC.

Area of Science:

• Robotics and Autonomous Systems
• Computer Vision
• Optical Communications

Background:

• Accurate vehicle-to-vehicle (V2V) positioning is crucial for advanced driver-assistance systems and autonomous driving.
• Monocular camera-based positioning methods often suffer from accuracy limitations due to an unfixed baseline.

Purpose of the Study:

• To develop an effective V2V positioning method utilizing a monocular camera and visible light communication (VLC).
• To enhance positioning accuracy by addressing the challenge of an unfixed baseline in monocular systems.

Main Methods:

• Proposed a V2V positioning method using a monocular camera integrated with visible light communication (VLC).
• Employed the known length of vehicle taillights as a fixed baseline to improve monocular positioning accuracy.
• Integrated a Kalman filter (KF) to mitigate random errors and refine vehicle position estimation.

Main Results:

• Experimental validation was conducted using a controllable mobile platform.
• The system demonstrated centimeter-level positioning accuracy under varying inter-vehicle distances and relative speeds.
• The proposed method effectively enhances the accuracy of V2V positioning using VLC.

Conclusions:

• The developed V2V positioning method based on monocular camera and VLC is effective and feasible.
• Utilizing taillight length as a fixed baseline and KF significantly improves positioning accuracy.
• Centimeter-level accuracy is achievable, paving the way for enhanced V2V communication and safety systems.