WebGear_RTC API Usage Examples:¶

Requirements¶

Installation with Asyncio Support¶

WebGear_RTC API is the part of asyncio package of VidGear, thereby you need to install VidGear with asyncio support as follows:

pip install vidgear[asyncio]

Aiortc¶

Must Required with WebGear_RTC API. You can easily install it via pip:

Microsoft Visual C++ 14.0 is required.

Installing aiortc on windows requires Microsoft Build Tools for Visual C++ libraries installed. You can easily fix this error by installing any ONE of these choices:

While the error is calling for VC++ 14.0 - but newer versions of Visual C++ libraries works as well.

Microsoft Build Tools for Visual Studio.
Alternative link to Microsoft Build Tools for Visual Studio.
Offline installer: vs_buildtools.exe

Afterwards, Select: Workloads → Desktop development with C++, then for Individual Components, select only:

Windows 10 SDK
C++ x64/x86 build tools

Finally, proceed installing aiortc via pip.

  pip install aiortc

ASGI Server¶

You'll also need to install an ASGI Server to run following WebGear_RTC usage examples, and by default WebGear_RTC ships the state-of-the-art uvicorn Server. But you can also use other ASGI server such as daphne, or hypercorn with it.

Bare-Minimum Usage¶

Let's implement a Bare-Minimum usage example:

Running Programmatically ¶

You can access and run WebGear_RTC VideoStreamer Server programmatically in your python script in just a few lines of code, as follows:

For accessing WebGear_RTC on different Client Devices on the network, use "0.0.0.0" as host value instead of "localhost" on Host Machine. More information can be found here ➶

We are using frame_size_reduction attribute for frame size reduction (in percentage) to be streamed with its options dictionary parameter to cope with performance-throttling in this example.

# import required libraries
import uvicorn
from vidgear.gears.asyncio import WebGear_RTC

# various performance tweaks
options = {
    "frame_size_reduction": 25,
}

# initialize WebGear_RTC app
web = WebGear_RTC(source="foo.mp4", logging=True, **options)

# run this app on Uvicorn server at address http://localhost:8000/
uvicorn.run(web(), host="localhost", port=8000)

# close app safely
web.shutdown()

which can be accessed on any browser on your machine at http://localhost:8000/.

Running from Terminal ¶

You can also access and run WebGear_RTC Server directly from the terminal commandline. The following command will run a WebGear_RTC VideoStreamer server at http://localhost:8000/:

Make sure your PYTHON_PATH is set to python 3.7+ versions only.

If you're using --options/-op flag, then kindly wrap your dictionary value in single '' quotes.

python3 -m vidgear.gears.asyncio --mode webrtc --source test.avi --logging --options '{"frame_size_reduction": 50, "frame_jpeg_quality": 80, "frame_jpeg_optimize": True, "frame_jpeg_progressive": False}'

which can also be accessed on any browser on the network at http://localhost:8000/.

Advanced Usage from Terminal

You can run python3 -m vidgear.gears.asyncio -h help command to see all the advanced settings, as follows:

usage: python -m vidgear.gears.asyncio [-h] [-m MODE] [-s SOURCE] [-a API] [-S] [-b BACKEND]
                [-cs COLORSPACE] [-cn CAMERA_NUM] [-r RESOLUTION] [-f FRAMERATE]
                [-yt] [-sd SOURCE_DEMUXER] [-ff FRAME_FORMAT] [-cf CUSTOM_FFMPEG]
                [-td TIME_DELAY] [-ip IPADDRESS] [-pt PORT] [-l] [-op OPTIONS]

Runs WebGear/WebGear_RTC Video Server through terminal.

options:
-h, --help            show this help message and exit
-m {mjpeg,webrtc}, --mode {mjpeg,webrtc}
                        Whether to use "MJPEG" or "WebRTC" mode for streaming.
-a {camgear,pigear,ffgear}, --api {camgear,pigear,ffgear}
                        Selects the capture backend for VideoGear. Choices: camgear, pigear, ffgear. Default: camgear.
-ep, --enablePiCamera
                        [DEPRECATED] Use `--api pigear` instead. Sets the flag to access PiGear API.
-S, --stabilize       Enables real-time video stabilization.
-s SOURCE, --source SOURCE
                        Path to input source (device index, filepath, URL, or glob pattern).
-yt, --stream_mode    Enables YouTube/yt_dlp Stream Mode in CamGear/FFGear API.
-b BACKEND, --backend BACKEND
                        Sets the backend of the video source in CamGear API (e.g. cv2.CAP_DSHOW).
-sd SOURCE_DEMUXER, --source_demuxer SOURCE_DEMUXER
                        [FFGear only] FFmpeg demuxer for the source (e.g. "v4l2", "dshow"). Default: auto-detect.
-ff FRAME_FORMAT, --frame_format FRAME_FORMAT
                        [FFGear only] Pixel format for decoded frames (e.g. "bgr24", "gray"). Default: bgr24.
-cf CUSTOM_FFMPEG, --custom_ffmpeg CUSTOM_FFMPEG
                        [FFGear only] Path to a custom FFmpeg executable. Default: use PATH.
-cn CAMERA_NUM, --camera_num CAMERA_NUM
                        [PiGear only] Sets the camera module index.
-r RESOLUTION, --resolution RESOLUTION
                        [PiGear only] Sets the resolution (width,height) for the camera module.
-f FRAMERATE, --framerate FRAMERATE
                        [PiGear only] Sets the framerate for the camera module.
-cs COLORSPACE, --colorspace COLORSPACE
                        Sets the colorspace of the output video stream.
-td TIME_DELAY, --time_delay TIME_DELAY
                        Sets the time delay (in seconds) before start reading the frames.
-ip IPADDRESS, --ipaddress IPADDRESS
                        Uvicorn binds the socket to this ipaddress.
-pt PORT, --port PORT
                        Uvicorn binds the socket to this port.
-l, --logging         Enables/disables error logging, essential for debugging.
-op OPTIONS, --options OPTIONS
                        Sets the parameters supported by APIs (whichever being accessed) to the input videostream. Wrap your dict value in
                        single or double quotes.