• Install Homebrew

  • /usr/bin/ruby -e "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/master/install)"
    

  • vim ~/.bash_profile
    export PATH=/usr/local/bin:$PATH
    

    source ~/.bash_profile
    

• Install python

  • brew install python python3
    cd /usr/local/bin
    ln -s ../Cellar/python/2.7.14/bin/python2 python
    

• Install ffmpeg

  • brew install ffmpeg \
      --with-tools \
      --with-fdk-aac \
      --with-freetype \
     --with-fontconfig \
     --with-libass \
     --with-libvorbis \
     --with-libvpx \
     --with-opus \
     --with-x265
    

• Install opencv

  • brew tap homebrew/science
    brew install opencv3 --with-contrib --with-python3
    

• Install boto3, watchdog

  •   pip2 install boto3 watchdog
    

  or python -m pip install boto3 watchdog

Clone this repo:

git clone [email protected]:imyoungyang/video-streaming.git

Create an IAM service role to give Rekognition Video access to your Kinesis video streams and your Kinesis data streams. SNS Topic to recieve the reginition name.

python iam-role-helper.py --create
python sns-helper.py --create

python collection-helper.py --create

python index_faces.py ./young-yang.jpg Young

python video-stream-helper.py --create

python data-stream-helper.py --create

• run command to create a stream processor.

  python rekognition-process.py --create

• run command to check process status

  aws rekognition describe-stream-processor --name appStreamProcessor-videoFaceRek

• run command python rekognition-process.py --start to start the process

• Open terminal and exeucte the upload to kinesis videos

  • python watch_for_changes.py

• Execute face detection in another terminal

  • python face-detection-multi-files.py

• run command python get-rekognition-result.py

• AWS developer guide: recognize faces in a video stream