A Probability-Based Algorithm Using Image Sensors to Track the LED in a Vehicle Visible Light Communication System.

Area of Science:

• Computer Vision
• Optical Engineering
• Automotive Technology

Background:

• Visible Light Communication (VLC) systems utilize vehicle LED lights for data transmission.
• Traditional LED detection methods struggle with motion blur caused by vehicle movement.
• High-speed vehicle motion significantly degrades image quality, hindering LED detection and data extraction.

Purpose of the Study:

• To develop a robust probability-based algorithm for tracking vehicle LED lights in VLC systems.
• To overcome the limitations of traditional pixel intensity-based methods in the presence of motion blur.
• To enhance the reliability of LED detection and data extraction under dynamic driving conditions.

Main Methods:

• A probability-based algorithm integrating pixel intensity, optical flow, and statistical information from previous frames.
• Calculation of conditional probability for each pixel belonging to an LED.
• Determination of LED position based on the calculated probabilities.
• Simulations incorporating real-world scenarios like changing LED positions and motion blur.

Main Results:

• The proposed algorithm demonstrates improved accuracy in detecting LED positions compared to traditional methods.
• Effective mitigation of challenges posed by motion blur in high-speed vehicle scenarios.
• Successful simulation of various real-world conditions, validating algorithm robustness.

Conclusions:

• The probability-based algorithm offers a significant advancement in tracking vehicle LED lights for VLC.
• This method enhances the reliability of data extraction from blurred images in dynamic environments.
• The findings support the potential for more robust and efficient vehicle-to-vehicle communication using visible light.