Overview
Organizations often require efficient indoor tracking systems to monitor and count individuals in specific settings, such as office premises, shopping malls, or other public spaces. Travancore Analytics has developed an advanced AI based real time people tracking software and in/out counting solution leveraging Yolo v8. This system accurately tracks individuals entering and exiting designated areas for security, management, and resource allocation.
- Cost-effective indoor tracking solution
- Accurate counting of individuals entering and exiting a designated area
- Seamless integration with multiple CCTV cameras
- Real-time monitoring capabilities for actionable insights
Case
It is important for organizations to effectively regulate the flow of people to ensure security, gain valuable insights, and improve customer experience. The meticulous tracking and counting of individuals entering and exiting premises are essential for various operational needs. Sophisticated AI based real time people tracking software can identify unauthorized access to high-security settings such as offices and shopping centers and enhance safety measures in real time.
Real-time occupancy data helps authorities make informed decisions for better emergency crisis management. In the retail sector, it facilitates customer flow analysis during peak hours, providing invaluable insights for optimizing staffing and enhancing customer satisfaction. This software is also essential for crowd management at concerts and sporting events, ensuring a smooth and enjoyable experience for everyone.
Challenges
To develop software with this capability, we encountered two main challenges. One was obtaining suitable footage from the cameras to accurately count people. We addressed this by creating a system that could integrate footage from multiple cameras in real-time, considering factors such as camera angle, resolution, distance, and the number of cameras.The second challenge was selecting a model capable of processing the footage in real-time. After thorough research, we chose YOLO v8 as the foundation for our software due to its smooth and fast real-time processing capabilities.
Solution
An advanced YOLO v8 (You Only Look Once version 8) person-tracking software was developed to track and count people in real time. This tool offers several advantages such as improved detection, real-time performance, efficiency, lightweight design, and better feature extraction.To ensure the software functions properly, we have installed high-definition CCTV cameras at key points where people tracking and counting are necessary. The footage from these cameras is fed into the YOLO v8-powered software.
Each person identified is displayed as a distinct object with a bounding box, making it easy to differentiate them from other objects in the frame. Once a person is identified, the algorithm assigns a unique ID and continuously tracks them. The software can log all this data, including timestamps, unique person IDs, and other relevant metadata, to a database.
Reports can be generated to show the total number of people enter/exit a premise, peak traffic times, and average dwell time (how long people stay in an area), among other metrics. The system is designed to accommodate additional features in the future, such as face recognition and heat map generation.
Impact
Our AI based real time people tracking software and counting solution will empower organizations with precise, cost-effective monitoring, enhancing security, operational efficiency, and customer experience. The system’s seamless integration with CCTV infrastructure provides actionable, real-time data, supporting better decision-making in crisis management, optimizing staff allocation during peak times, and ensuring smooth crowd flow in public events. Its adaptability allows for future enhancements like face recognition and heat map generation, making it a valuable tool for dynamic, data-driven environments across various sectors.