Doge Streamer 🦊

doge-streamer is a tool for Doge live streaming, which you can find at:

• YouTube, or
• Twitch

This tool is based on OpenCV and FFmpeg.

OpenCV is used for capturing images from cameras, processing video, performing Doge detection, image processing, and overlaying text/images.

FFmpeg is used to encode the stream with x264, AAC, and flv. The stream is sent out to castr.io using RTMP.

Features:

• background subtraction based motion detection (using BackgroundSubtractorMOG2)
• automatically switches to active camera
• displays an image when cameras are inactive
• stream can be controlled remotely using doge-stream-helper
• sound from one audio source

TODO:

• better audio input handling
• add delightful background music
• train a TensorFlow model to detect Doge? perhaps use a combination of motion detection and an RNN model?

Quickstart

Install dependencies

On macOS:

$ brew install cmake ffmpeg opencv
...

On Debian based distros:

$ apt-get install -yqq \
    cmake             \
    build-essential   \
    libavcodec-dev    \
    libavdevice-dev   \
    libopencv-dev     \
    libavformat-dev   \
    libavutil-dev     \
    libswresample-dev \
    libswscale-dev    \
    libopencv-dev
...

Compile

To compile source code just run:

$ mkdir build
$ cd build
$ cmake .. -DCMAKE_BUILD_TYPE=Debug
$ make -j4

Run RTMP Server

$ docker run -p 1935:1935 tiangolo/nginx-rtmp
...

Run doge-streamer

$ ./doge-streamer

Watch & Enjoy

Use VLC or ffplay to connect to live video stream:

$ ffplay -sync ext rtmp://localhost/live/stream
...