I am currently doing a motion detection project that records down video when motion is detected. Right now there is no error when recording the video but when i check on the video, it is 0 bytes. Any help would be very much appreciated.
This is my code:
camera = cv2.VideoCapture(0)
fourcc = cv2.VideoWriter_fourcc(*'XVID')
out = cv2.VideoWriter('output.avi',fourcc, 20.0, (640, 480))
The problem comes when your input frame size does not match the output video.
out = cv2.VideoWriter('output.avi',fourcc, 20.0, (640, 480))
Here you are promissing that your output video is 640,480 and it depends on your input source (if you don't resize it)
You can either hard code it (checking the frame size of the input video or stream source) or using the following code:
w = cap.get(cv2.CAP_PROP_FRAME_WIDTH);
h = cap.get(cv2.CAP_PROP_FRAME_HEIGHT);
fourcc = cv2.VideoWriter_fourcc(*'DIVX')
out = cv2.VideoWriter('output.mp4',fourcc, 15.0, (int(w),int(h)))
My suggestion is grabbing a frame outside the while loop and declaring the VideoWritter there with the frame width and height. Also, try changing the codec from XVID to DIVX or MJPG if that does not work.
A mismatch between camera's color sensors and what you specify in the function can also cause this. It probably extends to any incorrect parameter in the writer call.
out = cv2.VideoWriter('output.mp4', fourcc, 20.0, (width, height), isColor=False)
isColor defaults to True; ran into this when trying to capture from greyscale Basler cameras
In my case I was saving images to a video, image_data.shape gave me (1080, 1920, 4), which I had to change to (1080, 1920, 3) using
image_data = image_data[:, :, :3]
MP4 video format does not support an alpha channel (transparency) by default.
Not directly related to the original question but thought I might put this here since I could not find this solution anywhere.
opencv_version -vshow?