Recherche avancée

Sur d’autres sites (4424)

• OpenCV's VideoCapture::open Video Source Dialog

  13 novembre 2015, par swtdrgn

  In my current project, when I call VideoCapture::open(camera device index) and the camera is in used by another program, it shows a Video Source dialog and returns true when I select a device that is already in use.

  However, in my [previous] experiment project, when I called VideoCapture::open(camera device index), it doesn’t show this dialog.

  I want to know what is causing the Video Source dialog to show and the program to behave differently from the experimental project.

  This is the source code to the experiment project :

  int main (int argc, char *argv[])
  
  {
  
  
  
      //vars
  
      time_duration td, td1;
  
      ptime nextFrameTimestamp, currentFrameTimestamp, initialLoopTimestamp, finalLoopTimestamp;
  
      int delayFound = 0;
  
      int totalDelay= 0;
  
  
  
      // initialize capture on default source
  
      VideoCapture capture;
  
      std::cout << "capture.open(0): " << capture.open(0) << std::endl;
  
      std::cout << "NOOO" << std::endl;
  
      namedWindow("video", 1);
  
  
  
      // set framerate to record and capture at
  
      int framerate = 15;
  
  
  
      // Get the properties from the camera
  
      double width = capture.get(CV_CAP_PROP_FRAME_WIDTH);
  
      double height = capture.get(CV_CAP_PROP_FRAME_HEIGHT);
  
  
  
      // print camera frame size
  
      //cout << "Camera properties\n";
  
      //cout << "width = " << width << endl <<"height = "<< height << endl;
  
  
  
      // Create a matrix to keep the retrieved frame
  
      Mat frame;
  
  
  
      // Create the video writer
  
      VideoWriter video("capture.avi",0, framerate, cvSize((int)width,(int)height) );
  
  
  
      // initialize initial timestamps
  
      nextFrameTimestamp = microsec_clock::local_time();
  
      currentFrameTimestamp = nextFrameTimestamp;
  
      td = (currentFrameTimestamp - nextFrameTimestamp);
  
  
  
      // start thread to begin capture and populate Mat frame
  
      boost::thread captureThread(captureFunc, &frame, &capture);
  
      // loop infinitely
  
      for(bool q=true;q;)
  
      {
  
          if(frame.empty()){continue;}
  
          //if(cvWaitKey( 5 ) == 'q'){ q=false; }
  
          // wait for X microseconds until 1second/framerate time has passed after previous frame write
  
          while(td.total_microseconds() < 1000000/framerate){
  
              //determine current elapsed time
  
              currentFrameTimestamp = microsec_clock::local_time();
  
              td = (currentFrameTimestamp - nextFrameTimestamp);
  
              if(cvWaitKey( 5 ) == 'q'){
  
                  std::cout << "B" << std::endl;
  
                  q=false;
  
                  boost::posix_time::time_duration timeout = boost::posix_time::milliseconds(0);
  
                  captureThread.timed_join(timeout);
  
                  break;
  
              }
  
          }
  
  
  
          // determine time at start of write
  
          initialLoopTimestamp = microsec_clock::local_time();
  
  
  
          // Save frame to video
  
          video << frame;
  
          imshow("video", frame);
  
  
  
          //write previous and current frame timestamp to console
  
          cout << nextFrameTimestamp << " " << currentFrameTimestamp << " ";
  
  
  
          // add 1second/framerate time for next loop pause
  
          nextFrameTimestamp = nextFrameTimestamp + microsec(1000000/framerate);
  
  
  
          // reset time_duration so while loop engages
  
          td = (currentFrameTimestamp - nextFrameTimestamp);
  
  
  
          //determine and print out delay in ms, should be less than 1000/FPS
  
          //occasionally, if delay is larger than said value, correction will occur
  
          //if delay is consistently larger than said value, then CPU is not powerful
  
          // enough to capture/decompress/record/compress that fast.
  
          finalLoopTimestamp = microsec_clock::local_time();
  
          td1 = (finalLoopTimestamp - initialLoopTimestamp);
  
          delayFound = td1.total_milliseconds();
  
          cout << delayFound << endl;
  
  
  
          //output will be in following format
  
          //[TIMESTAMP OF PREVIOUS FRAME] [TIMESTAMP OF NEW FRAME] [TIME DELAY OF WRITING]
  
          if(!q || cvWaitKey( 5 ) == 'q'){
  
              std::cout << "C" << std::endl;
  
              q=false;
  
              boost::posix_time::time_duration timeout = boost::posix_time::milliseconds(0);
  
              captureThread.timed_join(timeout);
  
              break;
  
          }
  
      }
  
  
  
      // Exit
  
      return 0;
  
  }

  

• How Opensl es on android control the audio speed ?

  6 décembre 2013, par user1882379

  I use NDK+FFMPEG on android to decode an video file, and transfer the decoded pcm data to opensles for play audio .I register the opengles for the pcm format：

    SLDataFormat_PCM format_pcm;
  
      format_pcm.formatType = SL_DATAFORMAT_PCM;
  
      format_pcm.numChannels = channel;
  
      format_pcm.samplesPerSec = rate * 1000;
  
      format_pcm.bitsPerSample = SL_PCMSAMPLEFORMAT_FIXED_16;
  
      format_pcm.containerSize = SL_PCMSAMPLEFORMAT_FIXED_16;

  also I convert the decoded pcm data to AV_SAMPLE_FMT_S16 format using swr_convert(...), but the play audio speed is very fast. but if you use sdl , the play speed is normal. It would be sdl can control the play audio speed, but opensles can not. I do not know whether it is my mistake using opensles or opensles can not control audio speed(need application to control)？

• How to add network caching to ffmpeg

  10 novembre 2018, par JohnnyF

  I have a network with random latency of 200-400ms per msg

  when i use VLC i add this command to the reciever node

  --network-caching=1000

  this adds 1 sec delay on the receiver, that fixes the problem with the network random delays

  how can i do it in FFMPEG (google finds only solutions of LOW delays)