Improving Autonomous Vehicle Perception through Evaluating LiDAR Capabilities and Handheld Retroreflectivity...

Area of Science:

• Road safety engineering
• Transportation technology
• Optical measurement

Background:

• Traffic signs are crucial for road safety, especially with the rise of autonomous vehicles (AVs).
• Current traffic sign evaluation methods require advancement to ensure optimal visibility and communication.
• Understanding the relationship between physical sign properties and sensor data is vital for future transportation systems.

Purpose of the Study:

• To analyze the correlation between traffic sign retroreflectivity and LiDAR intensity.
• To evaluate the effectiveness of LiDAR technology in assessing traffic sign visibility.
• To identify factors influencing traffic sign performance and compliance.

Main Methods:

• Utilized Python 3.10 for statistical analysis of retroreflectivity coefficients and LiDAR intensity data.
• Employed two LiDAR configurations (2LRLiDAR and 1CLiDAR) for data collection.
• Focused analysis on RA1 and RA2 traffic sign classes.

Main Results:

• Observed variations in retroreflectivity compliance rates across different sign categories and colors.
• Identified distinct patterns in LiDAR intensity distributions, suggesting its potential for visibility assessment.
• Found limited direct correlations between retroreflectivity and LiDAR intensity, highlighting areas for further study.

Conclusions:

• Traffic sign design and production need targeted improvements based on performance variations.
• LiDAR technology shows promise for evaluating traffic sign visibility, but requires further standardization.
• Enhanced understanding of sign retroreflectivity and LiDAR interaction is key to improving road safety.