Distinguishing Traffic Queues and Counting People using Sensors
Author: Jimmy Book
Abstract
Public transportation stations often have multiple, closely situated queues. Current computer vision (CV) models struggle to accurately identify and count individuals in these queues. This poses a significant challenge for efficient people counting and queue management systems. This study explores the best solutions for queue management systems.
Motivation of Research
The precision of people counting, as a subsystem inside the TeamDD group project, is critical for improving public transportation efficiency and user experience. Accurately distinguishing between closely situated queues and counting individuals within these queues can provide valuable data for optimizing station operations.
Research Questions
How could we distinguish close queues and count people using sensors?
Research Approach
• Literature Review: Examining existing research and technologies related to sensor-based queue management and people counting.
• Experimental Evaluation: Conduct experiments to test different sensor technologies and models in distinguishing and counting individuals in close queues.
Results
