Mask and social distancing Detection

Task 1 Detecting Masks

1. Detect faces using cvlib.detect_faces
2. Apply model to classifiy if the face is masked or not

Task 2 Detections of social distancing

3. Detect bounding boxes of person using cvlib.detect_common_objects
4. calculate centroid for all boxes
5. feed these centroid to sklearn.DBSCAN to get the clusters.

Multithreading

Run both these tasks in 2 different threads.

Display processed frames on Browser(Flask)

use streaming to stream the final frame from the app to the browser

Realtime and Post Session Statistics of Detections

Realtime person, distance violators and un masked count display Post Session statistics of overall detection with time for analysis.