COVID Tools - Computer Vision Tools for COVID-19
Link to open source: https://github.com/Daksh777/COVID-Tools
We aim to solve the problem of manual mask and heart rate detection with the help of COVID Tools. By making mask and heartbeat detection touchless, we aim to reduce COVID spread.
COVID Tools can detect the heart rate of a person by just the live video of a person by using Eulerian Magnification Algorithm - data is collected from the forehead over time to estimate the user's heart rate. This is done by measuring the average optical intensity in the forehead location in the subimage's green channel because the forehead (and the cheeks) have the maximum amount of veins which makes it possible to see the colour of your face change as blood rushes in and out of your head. With good lighting and minimal noise due to motion, a stable heartbeat can be obtained in 10-15 seconds.
Along with that COVID Tools is fully capable to detect masks and eliminate manual mask checkers in hospitals, stores, malls, etc. preventing COVID spread even further. The 1st model was trained to detect faces and the 2nd model was trained to do binary classification using OpenCV. Overall, the model was trained on 800 images.
