13
votes

I have a video file (i.e. https://www.example.com/myvideo.mp4) and need to load it with OpenCV.

Doing the equivalent with an image is fairly trivial:

imgReq = requests.get("https://www.example.com/myimage.jpg")
imageBytes = np.asarray(bytearray(data), dtype=np.uint8)
loadedImage = cv2.imdecode(image, cv2.IMREAD_COLOR)

I would like to do something similar to the following (where loadedVideo will be similar to what OpenCV returns from cv2.VideoCapture):

videoReq = requests.get("https://www.example.com/myimage.mp4")
videoBytes = np.asarray(bytearray(data), dtype=np.uint8)
loadedVideo = cv2.videodecode(image, cv2.IMREAD_COLOR)

But cv2.videodecode does not exist. Any ideas?


Edit: Seeing as this may be a dead end with only OpenCV, I'm open for solutions that combine other imaging libraries before loading into OpenCV...if such a solution exists.

2
if you download the video to your hard disk and try VideoCapture there, does it work? - Micka
Yes it would, but I can't imagine that's the most efficient method as it introduces disk IO into the problem rather than keeping everything in memory. I'm looking for the lowest latency possible here. - abagshaw
no, it was just a test to reduce error sources (like mossing codecs) - Micka
Do you want to hold the entire video in memory or do you want it to act as a streaming video? - zindarod
I'd like to hold the entire video in memory. - abagshaw

2 Answers

20
votes

It seems that cv2.videocode is not a valid OpenCV API either in OpenCV 2.x or OpenCV 3.x.

Below is a sample code it works in OpenCV 3 which uses cv2.VideoCapture class.

import numpy as np
import cv2

# Open a sample video available in sample-videos
vcap = cv2.VideoCapture('https://www.sample-videos.com/video/mp4/720/big_buck_bunny_720p_2mb.mp4')
#if not vcap.isOpened():
#    print "File Cannot be Opened"

while(True):
    # Capture frame-by-frame
    ret, frame = vcap.read()
    #print cap.isOpened(), ret
    if frame is not None:
        # Display the resulting frame
        cv2.imshow('frame',frame)
        # Press q to close the video windows before it ends if you want
        if cv2.waitKey(22) & 0xFF == ord('q'):
            break
    else:
        print "Frame is None"
        break

# When everything done, release the capture
vcap.release()
cv2.destroyAllWindows()
print "Video stop"

You may check this Getting Started with Videos tutorial for more information.

Hope this help.

2
votes

You will have to read the video using VideoCapture. there is no other way around that for now. unless you define it yourself.

remember a video is a combination of images changing at defined frame rate.

So You can read each frame in a while loop. as you apply the imdecode function.

import numpy as np
import cv2

cap = cv2.VideoCapture('https://www.example.com/myimage.mp4')

while(cap.isOpened()):
    ret, image = cap.read()    
    loadedImage = cv2.imdecode(image, cv2.IMREAD_COLOR)
    cv2.imshow('frame',loadedImage)
    if cv2.waitKey(1) & 0xFF == ord('q'):
        break

cap.release()
cv2.destroyAllWindows()