Recherche avancée

Sur d’autres sites (9589)

• Inserting an image inside a video every few frames using ffmpeg

  31 juillet 2023, par Erez Hochman

  Can I use FFMPEG to insert an image every 20 frames in a video ? 
I'm trying to create a subliminal message experiment and I thought it would be an easy way to make it but I can't find anything online.

  



  I tried to make something myself and created a script that :
  
1.splits a file into audio and video files
  
2.splits the video into frames
  
3.overwrites every 20th image in the sequence with the message image
  
4.re-encoding the video
  
5.concatenating it with the original audio

  



  this works but it's way more disk space consuming to be comfortable, is there a better way to do this ?
  
any advice or thought would be happily welcome.

  


• Generating Video from Downloaded Images Using Fluent-FFmpeg : Issue with Multiple Image Inputs

  11 août 2023, par Pratham Bhagat

  I am having trouble creating Video from multiple images using fluent-ffmpeg in node.js.

  


  Here, I am getting the images from rquest body and downloading them in **temp **directory

  


     const imageUrls = req.body.imageUrls;
   const timeInBetween = parseFloat(req.query.time_in_between) || 1.0;

const tempDir = path.join(
      context.executionContext.functionDirectory,
      "temp"
    );

const downloadedImages = await Promise.all(
      imageUrls.map(async (imageUrl, index) => {
        try {
          const response = await axios.get(imageUrl, {
            responseType: "arraybuffer",
          });
          const imageName = `image_${index + 1}.png`;
          const imagePath = path.join(tempDir, imageName);
          await fs.writeFile(imagePath, response.data);
          return imagePath;
        } catch (error) {
          context.log(`Error downloading ${imageUrl}: ${error.message}`);
          return null;
        }
      })
    );


  


  I want to combine these images that are in downloadedImages array and create a video

  


  const outputVideoPath = path.join(tempDir, "output.mp4");

    let ffmpegCommand = ffmpeg();

    for (let i = 0; i < downloadedImages.length; i++) {
      context.log(downloadedImages.length);
      ffmpegCommand
        .input(downloadedImages[i])

        .inputOptions(["-framerate", `1/${timeInBetween}`])
        .inputFormat("image2")
        .videoCodec("libx264")
        .outputOptions(["-pix_fmt", "yuv420p"]);
    }

    ffmpegCommand
      .output(outputVideoPath)
      .on("end", () => {
        context.log("Video generation successful.");
        context.res = {
          status: 200,
          body: "Video generation and cleanup successful.",
        };
      })
      .on("error", (err) => {
        context.log.error("Error generating video:", err.message);
        context.res = {
          status: 500,
          body: "Error generating video: " + err.message,
        };
      })
      .run();


  


  By running it and giving value of "time_in_between" as 2 I get video of 2 seconds with a single image.

  


  

  • Utilized Fluent-FFmpeg library to generate a video from a list of downloaded images.

  


  • Expected the video to include all images, each displayed for a specified duration.

  


  • Tried mapping through the image paths and using chained inputs for each image.

  


  • Expected the video to have a sequence of images displayed.

  


  • Observed that the generated video only contained the first image and was of 0 seconds duration.

  


  


• hardware conversion of image pixel format in ffmpeg ?

  18 janvier 2024, par dongrixinyu

  I am trying to decode an online rtmp video stream into RGB format frames, and then encoding RGB frames into an online stream.

  


  Task

  


  Here is what I do now :

  


  


  decoding a video stream to get images(RGB) ---> ai model process ---> encoding frames(RGB) to form a video stream in H264

  


  


  My scheme

  


  All my code in written in C with FFmpeg dependencies. The detailed steps are :

  


  


  rtmp/rtsp video stream ---> AVPacket ---(nvidia cuda)---> AVFrame(nv12 pix fmt) ---> AVFrame(RGB pix fmt) ---> AI process.

  


  


  


  AVFrame(RGB pix fmt) ---> AVFrame(nv12 pix fmt) ---(nvidia cuda)---> AVPacket ---> rtmp/rtsp video stream

  


  


  Now, the decoding and encoding part are run on NVIDIA GPU, which is quite fast.

  


  But the conversion of pixel format between AV_PIX_FMT_NV12 and AV_PIX_FMT_RGB is run on CPU, which is astonishingly CPU-consuming cause the size of video frame is 2k.

  


  My question

  


  So, is there any off-the-shelf method to fulfill the conversion of image pixel format on GPU (especially via cuda) directly ?