[Python] Fast splitting of video into chunks of 'x' frames

I am trying to split a video into chunks of 1000 frames each (it has to be done at frame level). I am currently using the opencv library, but it is very slow. It takes half an hour to split an 1-hour video into these equally-sized chunks.

This is the function I am currently using:

def split_video(video_path, output_dir, frames_per_chunk=1000):
cap = cv2.VideoCapture(video_path)
fps = cap.get(cv2.CAP_PROP_FPS)
frame_count = 0
chunk_count = 0
frames = []

# Create output directory if it doesn't exist
if not os.path.exists(output_dir):
os.makedirs(output_dir)

frame_idx = 0
while cap.isOpened():
ret, frame = cap.read()
if not ret:
break

frames.append(frame)
frame_count += 1
# Save chunk after every 'frames_per_chunk' frames
if frame_count % frames_per_chunk == 0:
chunk_filename = f'video_chunk_{chunk_count}.mp4'
chunk_path = os.path.join(output_dir, chunk_filename)
save_frames_to_video(frames, chunk_path, fps)
frames = []
chunk_count += 1
frame_idx +=1

# Save any remaining frames
if frames:
chunk_filename = f'video_chunk_{chunk_count}.mp4'
chunk_path = os.path.join(output_dir, chunk_filename)
save_frames_to_video(frames, chunk_path, fps)

print(f'video split into {chunk_count + 1} chunks')
cap.release()
return


Are there any other methods or libraries in Python I can use that are more efficient at this? I have looked into the ffmpeg library, but that seems to handle well only splitting by duration instead of number frames.

Continue reading...