You could use FFMPEG as a separate process (simplest way) and let it decode first IDR for you. Here you have a class FFMPEG that has GetThumbnail() method, to it you pass address of video file, address of the JPEG image to be made, and resolution that you want the image to be:
using System.Diagnostics;
using System.Threading;
public class FFMPEG
{
Process ffmpeg;
public void exec(string input, string output, string parametri)
{
ffmpeg = new Process();
ffmpeg.StartInfo.Arguments = " -i " + input+ (parametri != null? " "+parametri:"")+" "+output;
ffmpeg.StartInfo.FileName = "utils/ffmpeg.exe";
ffmpeg.StartInfo.UseShellExecute = false;
ffmpeg.StartInfo.RedirectStandardOutput = true;
ffmpeg.StartInfo.RedirectStandardError = true;
ffmpeg.StartInfo.CreateNoWindow = true;
ffmpeg.Start();
ffmpeg.WaitForExit();
ffmpeg.Close();
}
public void GetThumbnail(string video, string jpg, string velicina)
{
if (velicina == null) velicina = "640x480";
exec(video, jpg, "-s "+velicina);
}
}
Use like this:
FFMPEG f = new FFMPEG();
f.GetThumbnail("videos/myvid.wmv", "images/thumb.jpg", "1200x223");
For this to work, you must have ffmpeg.exe in folder /utils, or change the code to locate ffmpeg.exe.
There are other ways to use FFMPEG in .NET, like .NET wrappers, you could google for them. They basically do the same thing here, only better. So if FFMPEG gets your job done, I'd recomend to use .NET wrapper.
与恶龙缠斗过久,自身亦成为恶龙;凝视深渊过久,深渊将回以凝视…