Recherche avancée

Sur d’autres sites (8079)

• How to handle queueing of video encoding during multiple video uploads ?

  6 mars 2016, par Yash Desai

  I am working on developing a video streaming site where users can upload videos to the site (multiple videos at once using the uploadify jquery plugin).

  Now, I am faced with the question of encoding the videos to FLV for streaming them online.

  When should the video encoding process take place ? Should it take place immediately after uploads have finished (i.e redirect the user to upload success page, and then start encoding in the background using exec command for ffmpeg ?) However, using this approach, how do i determine if the encoding has finished successfully ? What if users upload a corrupt video and ffmpeg fails to encode it ? How do i handle this in PHP ?

  How do i queue encoding of videos since multiple users can upload videos at the same ? Does FFMpeg has its own encoding queue ?

  I also read about gearman and message queueing options such as redis and AMQP in another related SO thread. Are these one of the potential solutions ?

  I would really appreciate if someone could give answers to my questions.

• Converting cv::Mat image from BGR to YUV using ffmpeg

  10 mars 2016, par bot1131357

  I am trying to convert BGR image into YUV420P, but when I try to view each of the YUV planes separately, this is what I see.

  Shouldn’t cv::Mat::data and AVFrame::data[0] be packed in the same way ? I should be able to do a direct memcpy. Am I missing something ?

  Any ideas ?

  Mat frame;
  
  VideoCapture cap;
  
  if(!cap.open(0)){
  
      return 0;
  
  }
  
  
  
  // capture a frame
  
  cap >> frame;
  
  if( frame.empty() ) return 0;
  
  
  
  cv::Size s = frame.size();
  
  int height = s.height;
  
  int width = s.width;
  
  
  
  // Creating two frames for conversion
  
  AVFrame *pFrameYUV =av_frame_alloc();
  
  AVFrame *pFrameBGR =av_frame_alloc();
  
  
  
  // Determine required buffer size and allocate buffer for YUV frame 
  
  int numBytesYUV=av_image_get_buffer_size(AV_PIX_FMT_YUV420P, width,
  
                  height,1);          
  
  
  
  // Assign image buffers
  
  avpicture_fill((AVPicture *)pFrameBGR, frame.data,      AV_PIX_FMT_BGR24,
  
      width, height);
  
  
  
  uint8_t* bufferYUV=(uint8_t *)av_malloc(numBytesYUV*sizeof(uint8_t));
  
  avpicture_fill((AVPicture *)pFrameYUV, bufferYUV, AV_PIX_FMT_YUV420P,
  
          width, height);
  
  
  
  // Initialise Software scaling context
  
  struct SwsContext *sws_ctx = sws_getContext(width,
  
      height,
  
      AV_PIX_FMT_BGR24,
  
      width,
  
      height,
  
      AV_PIX_FMT_YUV420P,
  
      SWS_BILINEAR,
  
      NULL,
  
      NULL,
  
      NULL
  
      );
  
  
  
  // Convert the image from its BGR to YUV
  
  sws_scale(sws_ctx, (uint8_t const * const *)pFrameBGR->data,
  
        pFrameYUV->linesize, 0, height,
  
        pFrameYUV->data, pFrameYUV->linesize);
  
  
  
  
  
  
  
  // Trying to see the different planes of YUV
  
  Mat MY = Mat(height, width, CV_8UC1);
  
  memcpy(MY.data,pFrameYUV->data[0], height*width);
  
    imshow("Test1", MY); // fail
  
  Mat MU = Mat(height/2, width/2, CV_8UC1);
  
  memcpy(MU.data,pFrameYUV->data[1], height*width/4);
  
    imshow("Test2", MU); // fail
  
  Mat MV = Mat(height/2, width/2, CV_8UC1);
  
  memcpy(MV.data,pFrameYUV->data[2], height*width/4);
  
    imshow("Test3", MV); // fail  
  
  
  
  waitKey(0); // Wait for a keystroke in the window

• convert CCTV .264 video to one of common formats (mp4, avi, etc.) via ffmpeg or other cmdline tool

  21 mars 2016, par yy502

  I’ve got a CCTV cam that produces .264 format video clips. These clips plays fine just like any other normal video recording that you would expect on portable devices, but only with it’s manufacture provided App. When played directly using VLC or mplayer, only gray blocks are visible in the picture. I doubt it is propitiatory encoding, but some kind of hardware encoded raw h264 format that I’m just lacking the right combination of arguments/options for playback or convert using ffmpeg. ffmpeg -i does report the basic metadata correctly, but also crazy amount of frame errors.... but I know the video can be played fine.

  The Android App has the following files in its lib folder :
  

  I understand these files are not all for video decoding but also some other feature in the app. I’m just hoping someone could maybe determine what extra lib or option is needed to convert it with ffmpeg. e.g. libh264ToRGB565.so could be useful maybe...?

  This is a screenshot of what to expect from the sample video.
  

  And here is the sample video clip (1.3M, 1280x720p) : http://146.185.145.75/vid.264
  (MD5 = 0ae871484b3832984f46e6820b21c673)

  Any suggestion is appreciated.