A High-Precision, Real-Time, and Robust Indoor Visible Light Positioning Method Based on Mean Shift Algorithm and...

Area of Science:

• Computer Vision
• Robotics
• Indoor Navigation Systems

Background:

• Visible light positioning (VLP) is crucial for indoor navigation.
• Existing VLP systems often neglect robustness and real-time capabilities, limiting practical applications.
• High positioning accuracy, robustness, and real-time performance are essential for VLP systems.

Purpose of the Study:

• To propose a novel VLP method addressing accuracy, robustness, and real-time constraints.
• To enhance VLP performance using image sensors and Light Emitting Diodes (LEDs).
• To integrate the mean shift (MS) algorithm and unscented Kalman filter (UKF) for superior VLP.

Main Methods:

• Utilized the mean shift (MS) algorithm for accurate and robust positioning.
• Integrated the unscented Kalman filter (UKF) for high-speed target tracking and improved accuracy during LED obstruction.
• Employed a previously developed LED-ID recognition algorithm for initial LED localization and system initialization.

Main Results:

• Achieved a positioning accuracy of 0.42 cm.
• Maintained an average processing time of 24.93 ms per frame, demonstrating real-time capability.
• Demonstrated robustness with maintained accuracy of 1.41 cm even when half of the LED was shielded.

Conclusions:

• The proposed VLP algorithm exhibits excellent accuracy, strong robustness, and good real-time performance.
• The integration of MS and UKF significantly enhances VLP system capabilities for practical indoor navigation.
• This method offers a reliable solution for indoor positioning challenges where accuracy and real-time operation are critical.