Live Detecting and Counting On webcam that how many people went in front from it

1

Hi. I got an project in which I have to detect faces and count how many people pass through the web cam. Please Help. Thank You.
I have this code but it only counts the live faces not the total faces

import dlib
import cv2
import numpy as np

videoCapture = cv2.VideoCapture(0)

detector = dlib.get_frontal_face_detector()
res = 0
while True:
ret, frame = videoCapture.read()
frame = cv2.flip(frame, 1)

gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
faces = detector(gray)

num = 0
l=0

for face in faces:
    l = len(faces)
    x,y = face.left(),face.top()
    hi,wi = face.right(),face.bottom()
    cv2.rectangle(frame,(x,y),(hi,wi),(0,0,255),2)
    num = num+1

    cv2.putText(frame, 'face'+str(num), (x-12,y-12), cv2.FONT_HERSHEY_SIMPLEX, 0.7, (0,255,0),2)
cv2.putText(frame, 'Live Face '+str(l), (20, 40), cv2.FONT_HERSHEY_SIMPLEX, 1, (0, 255, 255), 2)
cv2.imshow('frame', frame)
con = cv2.waitKey(1)
if con == ord('q'):
    res = l
    break

print(“result”,res)
videoCapture.release()
cv2.destroyAllWindows()

2

well, obviously, you need to keep track of the faces you’ve already seen before :wink:

one way to do this, would be to use face recognition, so you only count faces, that havent been seen before …

which means, you need to setup a database of face features to track between frames !

this would be a good starting point

ps: rather ditch dlib.get_frontal_face_detector(), it’s terrible slow !