Recherche avancée

Sur d’autres sites (10988)

• How to deal with every frame of a video by use ffmpeg ?

  7 février 2019, par candrwow

  I am using tensorflow to do video object segmentation,but now I only know how to do it on image(png,jpg),I want to do it on short time video(15 seconds),how to get frame by frame in sequence and processing between each frame ?

  now I split mp4 to pngs by ffmpeg,then do segamentation for every pngs,finally compound pngs to mp4 and delete every pngs.split and compound like this :

  ffmpeg -i video.mp4 -r 24 ./split/%03d.png
  
  ffmpeg -f image2 -i imgFilePath -r 24 videoDesPath

  but the solution is not good,it generate many imgs in disk and need more io operation,if process fail,many pngs may be unable to recycle,I want to find a solution like below :

  1. convert a video to stream(java,because I’m using on Android)
  2. read stream to get 1st frame,do object segamentation(this need 150ms 1s,must run in child thread) and write segamented frame to a new stream(called segamented stream).
  3. repeat the above steps,finally convert a segamented stream to video.

  can you teach me how to convert video to stream and get frame from stream ?Thank you !

• avformat/matroskaenc : Reset cur_master_element when discarding master

  16 juin 2022, par Andreas Rheinhardt

  avformat/matroskaenc : Reset cur_master_element when discarding master
  Before this patch the muxer writes an invalid file
  
  (namely one in which the Projection master is a child of
  
  the Colour element) if the following conditions are met :
  
  a) The stream contains AVMasteringDisplayMetadata without primaries
  
  and luminance (i.e. useless AVMasteringDisplayMetadata).
  
  b) The stream contains AV_PKT_DATA_SPHERICAL side data.
  
  c) All the colour elements of the stream are equal to default
  
  (i.e. unknown).
  
  Fortunately these conditions are very unlikely to be met.
  Signed-off-by : Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
  

  • [DH] libavformat/matroskaenc.c

• How to read audio in python using Librosa ?

  24 mai 2019, par Samual

  I am trying to read an audio file in Librosa but getting the following error

  FileNotFoundError: [WinError 2] The system cannot find the file specified

  It is mentioned that somewhere I need to install ffmpeg but it did not solve the problem. I have ffmpeg installed at ffmpeg in c:\programdata\anaconda3\lib\site-packages (1.4).

  Here is an example code which I am using to read sample audio file

  import librosa
  
  filename = librosa.util.example_audio_file()
  
  print(filename)
  
  y, sr = librosa.load(filename)

  It is throwing this error at line number 4

  C:\ProgramData\Anaconda3\lib\site-packages\librosa\util\example_data\Kevin_MacLeod_-_Vibe_Ace.ogg
  
  ---------------------------------------------------------------------------
  
  FileNotFoundError                         Traceback (most recent call last)
  
   in <module>
  
        1 filename = librosa.util.example_audio_file()
  
        2 print(filename)
  
  ----> 3 y, sr = librosa.load(filename)
  
  
  
  C:\ProgramData\Anaconda3\lib\site-packages\librosa\core\audio.py in load(path, sr, mono, offset, duration, dtype, res_type)
  
      117 
  
      118     y = []
  
  --> 119     with audioread.audio_open(os.path.realpath(path)) as input_file:
  
      120         sr_native = input_file.samplerate
  
      121         n_channels = input_file.channels
  
  
  
  C:\ProgramData\Anaconda3\lib\site-packages\audioread\__init__.py in audio_open(path, backends)
  
      105     """
  
      106     if backends is None:
  
  --> 107         backends = available_backends()
  
      108 
  
      109     for BackendClass in backends:
  
  
  
  C:\ProgramData\Anaconda3\lib\site-packages\audioread\__init__.py in available_backends()
  
       84 
  
       85     # FFmpeg.
  
  ---> 86     if ffdec.available():
  
       87         result.append(ffdec.FFmpegAudioFile)
  
       88 
  
  
  
  C:\ProgramData\Anaconda3\lib\site-packages\audioread\ffdec.py in available()
  
      106         stdout=subprocess.PIPE,
  
      107         stderr=subprocess.PIPE,
  
  --> 108         creationflags=PROC_FLAGS,
  
      109     )
  
      110     proc.wait()
  
  
  
  C:\ProgramData\Anaconda3\lib\site-packages\audioread\ffdec.py in popen_multiple(commands, command_args, *args, **kwargs)
  
       92         cmd = [command] + command_args
  
       93         try:
  
  ---> 94             return subprocess.Popen(cmd, *args, **kwargs)
  
       95         except OSError:
  
       96             if i == len(commands) - 1:
  
  
  
  C:\ProgramData\Anaconda3\lib\subprocess.py in __init__(self, args, bufsize, executable, stdin, stdout, stderr, preexec_fn, close_fds, shell, cwd, env, universal_newlines, startupinfo, creationflags, restore_signals, start_new_session, pass_fds, encoding, errors, text)
  
      767                                 c2pread, c2pwrite,
  
      768                                 errread, errwrite,
  
  --> 769                                 restore_signals, start_new_session)
  
      770         except:
  
      771             # Cleanup if the child failed starting.
  
  
  
  C:\ProgramData\Anaconda3\lib\subprocess.py in _execute_child(self, args, executable, preexec_fn, close_fds, pass_fds, cwd, env, startupinfo, creationflags, shell, p2cread, p2cwrite, c2pread, c2pwrite, errread, errwrite, unused_restore_signals, unused_start_new_session)
  
     1170                                          env,
  
     1171                                          os.fspath(cwd) if cwd is not None else None,
  
  -> 1172                                          startupinfo)
  
     1173             finally:
  
     1174                 # Child is launched. Close the parent's copy of those pipe
  
  
  
  FileNotFoundError: [WinError 2] The system cannot find the file specified
  
  </module>

  I not sure what could be wrong here. Librosa and ffmpeg both are installed but not working at all. Let me know if you need more info.