Python CV2 提取带字幕的视频帧

Question

I have a video file result.mp4 with subtitles. I want to extract frames, but I want subtitles to be on those frames too, subtitles are in file as a separate track.

Currently I use this code to extract:

vidcap = cv2.VideoCapture("result.mp4")
success,image = vidcap.read()
count = 0
while success:
  cv2.imwrite("frame%d.jpg" % count, image)     
  success,image = vidcap.read()
  print('Read a new frame: ', success)
  count += 1

which works fine, but as I said, does not include subtitles on the frames. I am able to extract frames with subtitles only if I hardcode subtitles into video which looks different and takes a lot of time.

Is there a way for opencv to extract frames with subs that are on separate track?

FYI: I use this FFMPEG command to merge .mp4 with .srt :

ffmpeg -i ourvid.mp4 -i ourvid.srt -c:s mov_text -c:v copy -c:a copy result.mp4 -f srt

which completes in around a second

if I hardcode it however:

ffmpeg -i ourvid.mp4 -vf "subtitles=ourvid.srt" result.mp4

it takes on average at least half of the length of the source video (in completion time)

Answer 1

Essentially, I think cv2 is not able to do so, but ffmpeg can do that, with a command like this:

ffmpeg -i ourvid.mp4 -vf select='between(n\,x\,y)',subtitles=ourvid.srt -q:v 2 frames%d.jpg

where x and y are start and end of range to extract frames so I just use, for example

os.system("ffmpeg -i ourvid.mp4 -vf select='between(n\,3901\,3901)',subtitles=ourvid.srt -q:v 2 frames%d.jpg")

to run it from my python script

not the nice way, but this is the only one I found so far