Opencv realizes simple face tracking

Face tracking is to let the computer find the position of the face in a picture and frame it . stay OpenCV Some classifiers are encapsulated in , These classifiers are cascaded together in a certain order to form a cascaded classifier . Can face 、 nose 、 Recognize some parts such as eyes .

Cascade classifier location

Install it on your own python Location find location

classifier

 OpenCV There are two steps to realize face tracking , Load cascade classifiers and use classifiers for image recognition .

'''
 Create cascading classifiers 
cascade= cv2.CascadeClassifier(filename)

filename:  Classifier address 
cascade:  Classifier object 
'''

'''
 Use cascaded classifiers to recognize images 
objects = cascade.detectMultiScale(image,scaleFactor,minNeighors,flags,minSize,maxSize)

cascade:  Created Cascading Classifier object 
image:  Images to be analyzed 
scaleFactor:  Optional parameters , Zoom ratio when scanning the image 
minNeighors:  Optional parameters , At least how many detection results can be reserved in each candidate area to determine the face , The greater the value , The smaller the analysis error 
flags:  Optional parameters , It is recommended to use the default value 
minSize:  Optional parameters , Minimum target size 
maxSize:  Optional parameters , Maximum target size 

 Return value :
objects:  An array of capture areas ,x、y、w、h, Abscissa 、 Ordinate 、 wide 、 high 
'''

Code

import cv2

video = cv2.VideoCapture(r"C:\Users\Liu\Videos\videos\demo.mp4")

#  Load cascade classifiers 
cascade = cv2.CascadeClassifier(r"F:\python3\Lib\site-packages\cv2\data\haarcascade_frontalface_default.xml")

while video.isOpened():
    retval,img = video.read()
    
    #  Use classifiers to analyze images 
    faces = cascade.detectMultiScale(img)
    
    for x,y,w,h in faces:
        cv2.rectangle(img,(x,y),(x+w,y+h),(0,0,255),3)
    cv2.imshow("Video",img)
    
    key = cv2.waitKey(25)
    if key == 32:
        cv2.waitKey()
        continue
    elif key == 27:
        break

video.release()
cv2.destroyAllWindows()

  result