Recherche avancée

Sur d’autres sites (6464)

• Manual encoding into MPEG-TS

  4 juillet 2014, par Lane

  SO...

  I am trying to take a H264 Annex B byte stream video and encode it into MPEG-TS in pure Java. My goals is to create a minimal MPEG-TS, Single Program, valid stream and to not include any timing information information (PCR, PTS, DTS).

  I am currently at the point where my generated file can be passed to ffmpeg (ffmpeg -i myVideo.ts) and ffmpeg reports...

  [NULL @ 0x7f8103022600] start time is not set in estimate_timings_from_pts
  
  
  
  Input #0, mpegts, from 'video.ts':
  
  Duration: N/A, bitrate: N/A
  
  Program 1 
  
    Stream #0:0[0x100]: Video: h264 (Main) ([27][0][0][0] / 0x001B), yuv420p(tv, bt709), 1280x720 [SAR 1:1 DAR 16:9], 29.97 fps, 29.97 tbr, 90k tbn, 59.94 tbc

  ...it seems like this warning for start time is not a big deal... and ffmpeg is unable to determine how long the video is. If I create another mpeg-ts file from my video file (ffmpeg -i myVideo.ts -vcodec copy validVideo.ts) and run ffmpeg -i validVideo.ts I get...

  Input #0, mpegts, from 'video2.ts':
  
  Duration: 00:00:11.61, start: 1.400000, bitrate: 3325 kb/s
  
  Program 1 
  
    Metadata:
  
      service_name    : Service01
  
      service_provider: FFmpeg
  
    Stream #0:0[0x100]: Video: h264 (Main) ([27][0][0][0] / 0x001B), yuv420p(tv, bt709), 1280x720 [SAR 1:1 DAR 16:9], 29.97 fps, 29.97 tbr, 90k tbn, 59.94 tbc

  ...so you can see the timing information and bitrate is there and so is the metadata.

  My H264 video is comprised of only I and P Frames (with the SPS and PPS preceding the I Frame of course) and the way that I am creating my MPEG-TS stream is...

  1. Write a single PAT at the beginning of the file
  2. Write a single PMT at the beginning of the file
  3. Create TS and PES packets from SPS, PPS and I Frame (AUD NALs too, if this is required ?)
  4. Create TS and PES packets from P Frame (again, AUD NALs too, if required)
  5. For the last payload of either an I Frame or P Frame, add filler bytes to an adaptation field to make sure it fits into a full TS packet
  6. Repeat 3-5 for the entire file

  ...my PAT looks like this...

  4740 0010 0000 b00d 0001 c100 0000 01f0
  
  002a b104 b2ff ffff ffff ffff ffff ffff
  
  ffff ffff ffff ffff ffff ffff ffff ffff
  
  ffff ffff ffff ffff ffff ffff ffff ffff
  
  ffff ffff ffff ffff ffff ffff ffff ffff
  
  ffff ffff ffff ffff ffff ffff ffff ffff
  
  ffff ffff ffff ffff ffff ffff ffff ffff
  
  ffff ffff ffff ffff ffff ffff ffff ffff
  
  ffff ffff ffff ffff ffff ffff ffff ffff
  
  ffff ffff ffff ffff ffff ffff ffff ffff
  
  ffff ffff ffff ffff ffff ffff ffff ffff
  
  ffff ffff ffff ffff ffff ffff

  ...and my PMT looks like this...

  4750 0010
  
  0002 b012 0001 c100 00ff fff0 001b e100
  
  f000 c15b 41e0 ffff ffff ffff ffff ffff
  
  ffff ffff ffff ffff ffff ffff ffff ffff
  
  ffff ffff ffff ffff ffff ffff ffff ffff
  
  ffff ffff ffff ffff ffff ffff ffff ffff
  
  ffff ffff ffff ffff ffff ffff ffff ffff
  
  ffff ffff ffff ffff ffff ffff ffff ffff
  
  ffff ffff ffff ffff ffff ffff ffff ffff
  
  ffff ffff ffff ffff ffff ffff ffff ffff
  
  ffff ffff ffff ffff ffff ffff ffff ffff
  
  ffff ffff ffff ffff ffff ffff ffff ffff
  
  ffff ffff ffff ffff

  ...notice after the c100 00, the "ff ff", f0... says that we are not using a PCR... Also notice that I have updated my CRC to reflect this change to the PMT. My first I Frame packet looks like...

  4741 0010 0000 01e0
  
  0000 8000 0000 0000 0109 f000 0000 0127
  
  4d40 288d 8d60 2802 dd80 b501 0101 4000
  
  00fa 4000 3a98 3a18 00b7 2000 3380 2ef2
  
  e343 0016 e400 0670 05de 5c16 345d c000
  
  0000 0128 ee3c 8000 0000 0165 8880 0020
  
  0000 4fe5 63b5 4e90 b11c 9f8f f891 10f3
  
  13b1 666b 9fc6 03e9 e321 36bf 1788 347b
  
  eb23 fc89 5772 6e2e 1714 96df ed16 9b30
  
  252d ceb7 07e9 a0c7 c6e7 9515 be87 2df1
  
  81f3 b9d2 ba5f 243e 2d5c cba2 8ca5 b798
  
  6bec 8c43 0b5d bbda bc5b 6e7c e15c 84e8
  
  2f13 be84 

  ...you’ll notice after the 01e0 0000, 8000 00 is the PES header extension where I specify no PTS / DTS and the remaining length is zero. My first P Frame packet looks like...

  4741 001d
  
  0000 01e0 0000 8000 0000 0000 0109 f000
  
  0000 0141 9a00 0200 0593 ff45 a7ae 1acd
  
  f2d7 f9ec 557f cdb6 ba38 60d6 a626 5edb
  
  4bb9 9783 89e2 d7e1 102e 4625 2fbf ce16
  
  f952 d8c9 f027 e55a 6b2a 81c3 48d4 6a45
  
  050a f355 fbec db01 6562 6405 04aa e011
  
  50ec 0b45 45e5 0df7 2fed a3f8 ac13 2e69
  
  6739 6d81 f13d 2455 e6ca 1c6b dc96 65d5
  
  3bad f250 7dab 42e4 7ba9 f564 ee61 29fb
  
  1b2c 974c 6924 1a1f 99ef 063c b99a c507
  
  8c22 b0f8 b14c 3e4d 01d0 6120 4e19 8725
  
  2fda 6550 f907 3f87

  ...and whenever an I Frame or P Frame is ending, I have a TS packet with an adaptation field like...

  4701 003c b000 ffff ffff ffff ffff ffff
  
  ffff ffff ffff ffff ffff ffff ffff ffff
  
  ffff ffff ffff ffff ffff ffff ffff ffff
  
  ffff ffff ffff ffff ffff ffff ffff ffff
  
  ffff ffff ffff ffff ffff ffff ffff ffff
  
  ffff ffff ffff ffff ffff ffff ffff ffff
  
  ffff ffff ffff ffff ffff ffff ffff ffff
  
  ffff ffff ffff ffff ffff ffff ffff ffff
  
  ffff ffff ffff ffff ffff ffff ffff ffff
  
  ffff ffff ffff ffff ffff ffff ffff ffff
  
  ffff ffff ffff ffff ffff ffff ffff ffff
  
  ffff ffff ffff ffff ffff ffff

  ...where the first b0 bytes are the adaptation field stuffing bytes and the remaining ones are the final bytes of the I or P Frame. So as you can tell I can use ffmpeg and pass it my file to create a valid movie in any format. However, I need the file I create to be in the proper format and I cannot quite figure out what the last piece I am missing is. Any ideas ?

• Convert opencv mat frame to ffmpeg AVFrame

  1er avril 2015, par Yoohoo

  I am currently work on a c++ real-time video transmission project, now I am using opencv to capture the webcam, and I want to convert the opencv Mat to ffmepg AVFrame to do the encoding and write into buffer. At the decoder side, read packet from buffer, using ffmpeg decode, and then convert ffmpeg AVFrame to opencv Mat again and play.

  Now I have finished the opencv capturing, and I can encode v4l2 source with ffmpeg, but what I want to do is replace the v4l2 source with opencv Mat. But it get error in follow code (I just show the part of the conversion) :

      Mat opencvin;                    //frame from webcam
  
      cap.read(opencvin);
  
  
  
      Mat* opencvframe;
  
      opencvframe = &opencvin;
  
      AVFrame ffmpegout;
  
      Size frameSize = opencvframe->size();
  
      AVCodec *encoder = avcodec_find_encoder(AV_CODEC_ID_H264);
  
      AVFormatContext* outContainer = avformat_alloc_context();
  
      AVStream *outStream = avformat_new_stream(outContainer, encoder);
  
      avcodec_get_context_defaults3(outStream->codec, encoder);
  
  
  
      outStream->codec->pix_fmt = AV_PIX_FMT_YUV420P;
  
      outStream->codec->width = opencvframe->cols;
  
      outStream->codec->height = opencvframe->rows;
  
      avpicture_fill((AVPicture*)&ffmpegout, opencvframe->data, PIX_FMT_BGR24, outStream->codec->width, outStream->codec->height);
  
      ffmpegout.width = frameSize.width;
  
      ffmpegout.height = frameSize.height;

  This is a code I borrow from Internet, it seems the frame have already be encoded during conversion, before I use

  static AVCodecContext *c= NULL;
  
  
  
  c = avcodec_alloc_context3(codec);
  
  if (!c) {
  
      fprintf(stderr, "Could not allocate video codec context\n");
  
      exit(1);
  
  }
  
  c->bit_rate = 400000;
  
  /* resolution must be a multiple of two */
  
  c->width = 640;
  
  c->height = 480;
  
  /* frames per second */
  
  c->time_base= (AVRational){1,25};
  
  c->gop_size = 10; /* emit one intra frame every ten frames */
  
  c->max_b_frames=15;
  
  c->pix_fmt = AV_PIX_FMT_YUV420P; 
  
  
  
  ret = avcodec_encode_video2(c, &pkt, &ffmpegout, &got_output);

  to encode the frame. And I get core dumped error if I continue encode the converted frame.

  I want encode the frame after conversion so that I can keep the data in pkt. How can I get a pure converted ffmpeg frame from opencv frame ?

  Or if the encode have already done during coversion, how can I get the output into pkt ?

• avcodec_open segmentation fault

  24 juin 2014, par user3715403

  Good afternoon,

  I got a copy of the ffmpeg’s api-example.c (https://www.ffmpeg.org/doxygen/0.6/api-example_8c-source.html). I tested the function video_encode_example() and got the code working as expected.

  I then re-factored this function into different class methods and now I get a seg fault with the call to avcodev_open(). I then changed the code to call the original video_encode_example() from a class method and the calls to avcodec... were successful.

  It appears that when called from a function (eg video_encode_example()), the behaviour is as expected, but avcodev_open fails when called from a class method. I traced the code and in both cases the values for the avcodev_open() arguments are assigned (and similar) pointer values. cgdb reports that the seg fault occurs in avcodev_open2()

  Any suggestions ?
  The code was compiled and tested on a Ubuntu 12.04 64 bit machine) using the libav package available thru apt-get (libavcodec.so.53.35.0)

  Regards,
  Daniel

  Added comment following the first response :

  From the link above, I copied the function video_encode_example() and placed the call inside a class constructor. This part is working.

  VideoEncoder::VideoEncoder( const std::string& aFileName)
  
      : mCodec( NULL)
  
      , mCodecContext( NULL)
  
      , picture (NULL)
  
      , mFileName( aFileName)
  
      , mFileHandler( NULL)
  
  {
  
      // must be called before using avcodec lib
  
      ::avcodec_init();
  
  
  
      // register all the codecs
  
      ::avcodec_register_all();
  
  
  
      video_encode_example( aFileName.c_str());
  
      return;
  
      ...
  
  }

  The re-factored part includes splitting the avcodev call (from the original video_encode_example() function) to different VideoEncoder methods. The constructor now looks like :

  VideoEncoder::VideoEncoder( const std::string& aFileName)
  
      : mCodec( NULL)
  
      , mCodecContext( NULL)
  
      , picture (NULL)
  
      , mFileName( aFileName)
  
      , mFileHandler( NULL)
  
  {
  
      // must be called before using avcodec lib
  
      ::avcodec_init();
  
  
  
      // register all the codecs
  
      ::avcodec_register_all();
  
  
  
      // find the mpeg1 video encoder
  
      mCodec = ::avcodec_find_encoder( CODEC_ID_MPEG1VIDEO);
  
      if (!mCodec) {
  
          fprintf( stderr, "codec not found\n");
  
          exit( 1);
  
      }
  
  
  
      mCodecContext = ::avcodec_alloc_context3( mCodec);
  
      picture = ::avcodec_alloc_frame();
  
  
  
      // put sample parameters
  
      mCodecContext->bit_rate = 400000;
  
  
  
      // frames per second
  
      mCodecContext->time_base = (AVRational){1,25};
  
      mCodecContext->gop_size = 10;               // emit one intra frame every ten frames
  
      mCodecContext->max_b_frames = 1;
  
      mCodecContext->pix_fmt = PIX_FMT_YUV420P;
  
  
  
      // open it
  
      // Initializes the AVCodecContext to use the given AVCodec. Prior to using this function the context has to be allocated.
  
      if (::avcodec_open( mCodecContext, mCodec) < 0) {
  
          fprintf(stderr, "could not open codec\n");
  
          exit( 1);
  
      }
  
  
  
      mFileHandler = fopen( mFileName.c_str(), "wb");
  
      if (!mFileHandler) {
  
          fprintf( stderr, "could not open %s\n", mFileName.c_str());
  
          exit( 1);
  
      }
  
  
  
  }

  The call to avcodec_open causes the seg fault. In addition, the fault occurs whether using avcodev_.. or ::avcodev_...