A Computer Vision based solution to check if people follow social distancing norms, wear a mask, or if they are roaming out!
All the necessary modules and the dependancies can be downloaded using pip installer
using the requirements.txt
file
Best suited to run the program on IDLE and Python 3.7 on a Windows system.
- Download the zip file of the project
- Extract the files into the same folder
- Run the
GUI-RUN THIS.py
file - The user-friendly menu window will pop up, use it navigate through various modules
Click the image to watch the demo of the entire project
This is how the menu page looks like, you can click the buttons for respective modules to work
Click the video below to see how to add a COVID patient into the database. The status can be set as positive or at risk.
Click the video below to see how to delete a record of a COVID patient. If the person recovers, we can then remove that person's record from the database.
Click the video below to see the demonstration of how the module classifies a person with mask or without mask. Can be implemented in places where we can automate surveillance of Face mask wearers.
Click the video below to see how the module recognizes faces, and marks them with names and their status Positive or At risk. Can be implemented for automating the surveillance of those patients who are roaming in public and can further spread the diseases. Also, maintains a database of the ID, Person's name, Date and Time when the person was last spotted.
You can even detect more than 1 patient in one frame, example below!
Click the video below to see how the module tracks people and marks them with red/green boxes to keep a track of the amount of social distance violations. Can be implemented for automating the surveillance rather than using actual police force. Also shows the necessary statistics.
Click the video below to see how RAKSHAK - Web based all in one app can be used for Face Mask Detection, Social Distancing Detector, tracking COVID positive patients using centralized cloud data
Built using
- Python 3
- TensorFlow
- Yolo v3
- Pretrained models
- HTML
- CSS
- Javascript
- Bootstrap
We would like to extend our heartful gratitude to the following people, without whom this project was not possible
- Adrian Rosebrock
- AI Comuter Systems
- mycodecamp.org
- Easy OpenCV