This machine learning project uses computer vision techniques to count the number of people entering and exiting a mall.
The system uses OpenCV and Python to detect and track people in video feeds from cameras placed at mall entrances and exits. The number of people entering and exiting is counted and displayed on a dashboard. The total count of people inside the mall is calculated and displayed. Alerts can be set to notify if the number of people exceeds a threshold. Data is stored in an Excel sheet for the admin to view and update.
The video feed from the cameras is broken down into frames. Each frame is processed using a pre-trained MobileNetSSD object detection model to detect people. Bounding boxes are drawn around each detected person. As people enter and exit frames, the counts are incremented or decremented accordingly. The total count is calculated and displayed on the dashboard. The data is stored in an Excel sheet when the system closes.
Programming Language: Python
Frameworks/Libraries:
OpenCV - For computer vision and image processing
Flask - For building the web application
MobileNetSSD - Pre-trained model for object detection
NumPy - For numerical operations
Pandas - For data manipulation and analysis
Database: SQLite
Deployment: Heroku
Version Control: Git/GitHub
IDE: Visual Studio Code
Design: HTML/CSS/JS
The detailed project report can be found here: [Report.pdf] (https://github.com/parthvasoya59/Crowd-Countings/blob/main/Report.pdf)