How to live stream to a website using cv2 library

asked 2020-03-10 12:59:21 -0600

So I have the following code:

import numpy as np
import cv2
fullbody_cascade = cv2.CascadeClassifier('haarcascade_fullbody.xml')
cap0 = cv2.VideoCapture('walking0.mp4')
cap1 = cv2.VideoCapture('walking1.mp4')

while True:
    ret0, frame0 = cap0.read()
    ret1, frame1 = cap1.read()
    gray0 = cv2.cvtColor(frame0, cv2.COLOR_BGR2GRAY)
    gray1 = cv2.cvtColor(frame1, cv2.COLOR_BGR2GRAY)
    fullbody0 = fullbody_cascade.detectMultiScale(gray0)
    fullbody1 = fullbody_cascade.detectMultiScale(gray1)
    for (x,y,w,h) in fullbody0:
        cv2.rectangle(frame0, (x,y), (x+w, y+h), (255,0,0), 2)
    for (x,y,w,h) in fullbody1:
        cv2.rectangle(frame1, (x,y), (x+w, y+h), (255,0,0), 2)
        #roi_gray = gray[y:y+h, x:x+w]
        #roi_color = img[y:y+h, x:x+w]

    cv2.imshow('cam0',frame0)
    cv2.imshow('cam1',frame1)
    k = cv2.waitKey(30) & 0xff
    if k == 27:
        break
cap.release()
cv2.destroyAllWindows()

And what this does is use a trained Haar Cascade of a person's fullbody and draws an outline around it and "cv2.imshow" is used to preview what is being seen. What I actually want to do is display it on a webpage instead, whether it be a ddns server, localhost or a domain, any is fine. I initially used Motion with my Raspberry Pi 3 to stream to localhost but it is not what I need. I've seen people use Python Flask but I do not think it is compatible with my application, I may be wrong. To explain the code more, it's processing 2 videos; "walking0.mp4" and "walking1.mp4" and outlining the humans in both videos at the same time. Does anyone have any suggestions on how I can modify this code to get my desired outcome?

edit retag flag offensive close merge delete

Comments

Raspberry pi or pc?

supra56 gravatar imagesupra56 ( 2020-03-11 04:43:11 -0600 )edit

Raspberry pi.

dunnoguy gravatar imagedunnoguy ( 2020-03-11 10:16:20 -0600 )edit