Social Distance Approximation on Public Transport Using Stereo Depth Camera and Passenger Pose Estimation

Area of Science:

• Public Health
• Transportation Engineering
• Computer Vision

Background:

• Pandemic necessitates balancing public transport operations with infection control.
• Understanding passenger behavior and environmental conditions is crucial for safety.
• Social distancing effectiveness on public transport requires quantitative analysis.

Purpose of the Study:

• To assess the feasibility of social distancing on public transport.
• To analyze passenger behavior and seating preferences in relation to social distancing.
• To evaluate the utility of computer vision for monitoring social distancing.

Main Methods:

• Deployment of a low-cost depth-sensing system on a public bus.
• Utilizing advanced computer vision techniques to measure social distancing.
• Analyzing passenger habits and seating preferences based on acquired data.

Main Results:

• Social distancing is unlikely beyond a 28% bus occupancy threshold.
• High probability (89%) of being within 1-2m of another passenger at 28% occupancy.
• Passenger preferences prioritize comfort and views over maximizing social distancing.

Conclusions:

• Current public transport design and passenger behavior limit effective social distancing.
• Computer vision offers a robust method for monitoring environmental conditions and passenger behavior.
• Findings inform strategies for safer public transport operations during health crises.