FFMPEG API - Retrieve device camera information - c++

I need to get the compatible framerate/resolution of the available cameras. How can this be done with the ffmpeg library? I've tried using the functions of avdevice but all of them seem to retrieve an error. I am not able to get a list of the available devices as well.
This is being done on a mac using avfoundation (and it will later be ported to windows with dshow).
Thank you for your time.

Try running
ffmpeg -f dshow -list_options true -i video="Integrated Camera"
replacing 'dshow' and "Integrated Camera" with whatever you have, depending on the platform. As above, you can get the name of the video device with
ffmpeg -f dshow -list_devices true -i x
You can then pipe the resulting output to a file using the > operator, or to a further command-line tool for processing with the | operator. For example,
ffmpeg -f dshow -list_options true -i video="Integrated Camera" > test.txt
or
ffmpeg -f dshow -list_options true -i video="Integrated Camera" | grep 'pixel_format'

Related

How to access framebuffer with FFMPEG?

I'm writing an application which need to capture screen. I've looked up for solution and internet says that FFMPEG could do it. But I can't find the way to do that IN CODE. FFMPEG documentation seems to be very poor.
Can anybody please tell me how do I access framebuffer raw data with FFMPEG?
FFmpeg supports input of rawframes throught stdin:
With the arg -f rawvideo ffmpeg will expect frames coming from stdin
ffmpeg -r 60 -f rawvideo -pix_fmt uyvy422 -s 1280x720 -i - -threads 0 -preset fast -y -pix_fmt yuv420p output.mp4
You can check this link, it has useful information.
In Qt, you would run a QProcess with ffmpeg with -f rawvideo and write to stdin with write() method.
This is roughly how to acomplish it:
QProcess* process;
process->start("ffmpeg.exe", args, QProcess::Unbuffered | QProcess::ReadWrite);
process->waitForStarted();
...
process->setProcessChannelMode(QProcess::ForwardedChannels);
videoFrame->GetBytes(&buffer);
process->write(buffer);

how to use a video stream as input in python/opencv programme

I am able to stream and receive webcam feed in two terminal via udp
command for streaming:
ffmpeg -i /dev/video0 -b 50k -r 20 -s 858x500 -f mpegts udp://127.0.0.1:2000
command for recieving:
ffplay udp://127.0.0.1:2000
Now i have to use this received video stream as input in python/opencv how can i do that.
I will be doing this using rtp and rstp as well.
But in case of rtsp it is essential to initiate the receiving terminal, but if I do that then port will become busy and my program will not be able to take the feed.How could it be resolved.
I am currently using opencv 2.4.13, python 2.7 in ubuntu 14.04
Check this tutorial, and use cv2.VideoCapture("udp://127.0.0.1:2000"). You will need to build opencv with FFmpeg so that it works.

ffmpeg doesn't work when called from c++ system

I have a c++ script that coneverts a series of jpg into a .mp4 video, the command i use is the folllowing:
std::system("ffmpeg -threads auto -y -framerate 1.74659 -i /mnt/ev_ramdsk/1/%05d-capture.jpg -vcodec libx264 -preset ultrafast /mnt/ev_ramdsk/1/video.mp4");
which produces a .mp4 video file like its supposed to except it cant be played from anywhere (tested in 2 computers and html5 video)
But, if from the same computer where the program runs, i do:
ffmpeg -threads auto -y -framerate 2 -i %05d-capture.jpg -vcodec libx264 -preset ultrafast video.mp4
from the command line, the output video plays wonderfully (except in vlc, for vlc i have to use -vcodec mpeg4)
What can possibly cause this behaviour?
could cp command corrupt the file? (ran after the mpeg to move it out of the ramfs)
EDIT:
As requested, i ran the whole set of commands one by one in the console exactly as the program do (the program logs every single command it runs, i just repeated them).
The commands are:
cp -r /var/cache/zoneminder/events/1/16/05/18/23/30/00/ /mnt/ev_ramdsk/1/
ffmpeg -threads auto -y -framerate 1.76729 -i /mnt/ev_ramdsk/1/%5d-capture.jpg -preset ultrafast /mnt/ev_ramdsk/1/video.mp4
cp /mnt/ev_ramdsk/1/video.mp4 /var/cache/evmanager/videos/1/2016_05_18_23_30_00_.mp4
The resulting .mp4 file can be played without any trouble. Also, is the only one with a preview image in the file explorer.
Thank you very much!
Solved it!
this was the winning answer. finally got it to work using:
std::system("ffmpeg -threads auto -y -r 1.74659 -i /mnt/ev_ramdsk/1/%05d-capture.jpg -px_fmt yuv420p -preset ultrafast -r 10 /mnt/ev_ramdsk/1/video.mp4");
Thank you very much!

FFMPEG command for audio file streaming

I am trying to stream an audio file in mp3 format using the FFMPEG library to a remote computer, located on the same LAN as the sender. The command i used to stream at the sender is given below:
ffmpeg -re -f mp3 -i sender.mp3 -ar 8000 -f mulaw -f rtp rtp://10.14.35.23:1234
I got the below command on FFMPEG documentation page that generates audio and streams it to port number 1234 on remote computer
ffmpeg -re -f lavfi -i aevalsrc="sin(400*2*PI*t)" -ar 8000 -f mulaw -f rtp rtp://10.14.35.23:1234
I thought i had made relevant changes to this so that the mp3 streaming command will work, but only to know encounter the error which reads
"Error while opening encoder for output stream #0:0 - maybe incorrect parameters such as bit_rate, rate, width or height"
Can anyone tell me what is the wrong parameter here and how to rectify it?
I could figure out the way to stream an audio file using FFMPEG. The command for the same is given below:
ffmpeg -re -f mp3 -i sender.mp3 -acodec libmp3lame -ab 128k -ac 2 -ar 44100 -f rtp rtp://10.14.35.23
Here the audio file 'sender.mp3' is located in the same folder as ffmpeg.exe. In case of a different folder, the full path should be mentioned in the command.

How to generate a thumbnail of flash movie (.flv)

I'm using ColdFusion and need to generate a thumbnail from a flash movie stored on the server. I have heard of ffMpeg but have no idea how to use it. (Once you put it on your server what's the next step?)
You can use cfexecute to run a command line on the CF server.
Karthik linked a blog post that suggests the following syntax for ffmpeg:
ffmpeg -itsoffset -4 -i test.avi
-vcodec mjpeg -vframes 1 -an -f rawvideo -s 320x240 test.jpg
So you could do something like this:
<cfexecute
name="c:\pathto\ffmpeg\ffmpeg.exe"
arguments="-itsoffset -4 -i #sourcevideo# -vcodec mjpeg -vframes 1 -an -f rawvideo -s 320x240 #thumbnaildestination" />
I haven't run ffmpeg like this and you'll likely need to experiment with the syntax to get a result you like, but once you do your workflow is pretty straightforward.
You may also run into issues executing fmpeg.exe depending on the user account your ColdFusion server instance is running as.
Documentation of FFMpeg: http://www.ffmpeg.org/documentation.html
You might want to check: http://blog.prashanthellina.com/2008/03/29/creating-video-thumbnails-using-ffmpeg/
http://www.flashcomguru.com/index.cfm/2006/4/25/ffmpegthumbs
with ColdFusion its not possible but check this: http://old.nabble.com/Create-a-thumbnail-image-from-.flv-video-file-once-uploaded-td22683497.html