ABSTRACT:
Crowd management and public safety are critical challenges in highly populated areas such as bus stands, railway platforms, and public gathering zones. Overcrowding in such environments can lead to accidents, panic situations, and inefficient movement of people if not monitored in real time. To address this issue, this project presents the design and implementation of a Raspberry Pi–based AI Crowd Density Monitoring and Alert System using computer vision and edge intelligence.
The proposed system utilizes a Raspberry Pi equipped with a CSI camera to capture live video streams, which are processed locally using a deep learning–based object detection model (YOLOv5) to detect and count the number of people present in the monitored area. Based on the detected count, crowd density is classified into three levels: Low, Medium, and High using predefined threshold logic. This classification enables timely decision-making without relying on cloud processing, ensuring low latency and enhanced data privacy.
Visual and audio alerts are integrated into the system using an LED indicator, buzzer, and LCD display. For medium crowd conditions, the LED provides a visual warning, while in high crowd situations, the system activates an audible buzzer and sends an automated alert message to authorized personnel using a communication service. Additionally, the system logs time-stamped crowd data, including count and density level, into a CSV file for further analysis, reporting, and planning purposes.
The proposed solution is low-cost, scalable, and suitable for continuous real-time operation. By leveraging edge AI and embedded systems, this project offers an effective and practical approach for crowd monitoring, safety enforcement, and smart infrastructure deployment in public spaces.


INTRODUCTION:
Rapid urbanization and population growth have significantly increased the density of people in public places such as bus stands, railway platforms, shopping complexes, and event venues. In such environments, unmanaged crowd accumulation can lead to serious safety risks including stampedes, delays, discomfort, and difficulty in emergency evacuation. Traditional crowd monitoring methods mainly rely on manual observation or conventional CCTV systems, which are often inefficient, labor-intensive, and unable to provide real-time actionable insights.
With recent advancements in artificial intelligence (AI) and embedded systems, automated crowd monitoring has become a feasible and effective solution. Computer vision–based approaches enable real-time detection and analysis of people using camera feeds, reducing the dependency on human supervision. However, most existing solutions are cloud-dependent, which introduces challenges such as high latency, increased operational cost, privacy concerns, and unreliable performance in low-network conditions.
To overcome these limitations, this project focuses on the development of an AI-based crowd density detection and alert system using Raspberry Pi. The system operates on the principle of edge computing, where all processing is performed locally on the device without requiring continuous internet connectivity. A deep learning object detection model is employed to detect and count people from live video streams captured by a camera. Based on the count, the crowd density is categorized into predefined levels, enabling timely alerts and visual indications.
The proposed system integrates both hardware and software components, including a camera, LED indicators, buzzer, LCD display, and alert mechanisms, to provide real-time feedback and warnings. Additionally, crowd data is logged for future analysis, making the system useful not only for immediate safety monitoring but also for long-term planning and management. Due to its low cost, scalability, and reliability, the system is well-suited for deployment in public transportation hubs and other crowded environments.

OBJECTIVES:
The main objective of this project is to design and develop an AI-based crowd density monitoring and alert system using Raspberry Pi for enhancing public safety in crowded environments. The specific objectives of the project are as follows:
1. To develop a real-time crowd detection system using a camera and deep learning techniques to accurately detect and count the number of people present in a monitored area.
2. To classify crowd density levels (Low, Medium, and High) based on the detected number of people using predefined threshold logic.
3. To implement an edge-based processing approach using Raspberry Pi in order to minimize latency, reduce dependency on cloud services, and ensure data privacy.
4. To provide visual and audio alerts using LED indicators and a buzzer to warn about increasing or dangerous crowd conditions in real time.
5. To display live crowd information such as people count and crowd status on an LCD for easy monitoring by authorities or operators.
6. To send automated alert notifications to responsible personnel when high crowd density is detected, enabling timely preventive action.
7. To record crowd data including time, people count, and density level into a CSV file for future analysis, reporting, and crowd management planning.
8. To design a low-cost and scalable solution that can be deployed in public places such as bus stands, railway platforms, and event venues.
9. To improve crowd safety and management efficiency by reducing the need for continuous manual monitoring.