VideoToSMI-Server

Create a smi file in Web based on the video

• Copyright (c) 2019 InfoLab (Donggun LEE)

• How to install

  • requirement

    # python 3.6 -- tensorflow or tensorflow-gpu
    pip install tensorflow-gpu==1.9.0 exifread>=2.1.2 piexif>=1.1.2 pillow>=6.0.0 matplotlib>=3.1.0 scikit-image>=0.15.0 IPython>=7.5.0 keras>=2.2.4 cython>=0.29.7 deepgeo VideoToSMI ConfigHelper

  • install

    pip install VideoToSMI-Server

  • other version

    # 0.0.3
    pip install VideoToSMI-Server==0.0.3
    # 0.0.4
    pip install VideoToSMI-Server==0.0.4

• How to use

  from VideoToSMIServer import Server, ServerConfig
  
  config = ServerConfig()
  config.MODEL_NAME = "mscoco"
  config.IP = "127.0.0.1"
  config.PORT=80
  config.MODEL_CONFIG_PATH = "D:/test/config.json"
  config.MODEL_ENGINE = "maskrcnn"
  config.FILTER = ['truck','car','bus']
  config.VIDEO_FOLDER = "D:/test/videotosmi/"
  server = Server(config)
  server.Run()
  
  # POST VIDEO FILE -> http://127.0.0.1:80/ -> Reulst SMI FILE

• How to Test

• Preview