About

CrowdMind AI is an intelligent system designed to improve safety in crowded environments such as stadiums, airports, festivals, and public events. The main question this project addresses is: How can artificial intelligence help detect crowd congestion early and guide people toward the safest evacuation routes during emergencies? Many crowd-related accidents occur because authorities cannot quickly analyze crowd density or predict where congestion will occur. This project aims to solve that problem using computer vision and AI-driven route optimization.

The success definition of the system is the ability to accurately detect the number of people in an area, estimate crowd density, identify potentially dangerous congestion zones, and recommend the safest and fastest evacuation routes. A successful system should provide real-time analysis, clear visualizations, and reliable exit recommendations that help prevent stampedes and improve emergency response.

The methodology involves several AI and data-processing steps. First, video input from cameras or recorded footage is processed using computer vision techniques. A person detection model identifies individuals in each frame and counts them. The system then divides the monitored area into grid cells to calculate people per square meter, generating a crowd density map. A heatmap visualization highlights safe and crowded areas.

Next, the system analyzes movement patterns and predicts where congestion may occur in the near future. Using graph-based algorithms such as Dijkstra’s algorithm, the environment is modeled as a network of paths and exits. The system calculates multiple evacuation routes and evaluates them based on factors like distance and congestion risk. The route with the lowest cost is recommended as the safest exit.

The findings from the prototype simulation show that the system can successfully analyze crowd conditions and determine optimal evacuation routes. In the test scenario, 260 people were detected in an 8000 m² area, indicating a low overall density. All zones were classified as safe, and the system recommended Exit A (North) as the most efficient evacuation path. The system also generated several visual outputs, including a crowd density heatmap, zone classification map, evacuation route visualization, and congestion prediction model.

Overall, CrowdMind AI demonstrates how AI-driven crowd monitoring and intelligent routing can enhance public safety and support smarter emergency management systems.