Add a Vehicle Model to a Database (Using Separate Steps)

This section describes how to create a new (empty) vehicle model, add training images, and then train Media Server, using separate actions. You can also add metadata to the vehicle model, but this is an optional step.

Alternatively, you can train Media Server to recognize a vehicle model by sending a single action. To do this, see Add a Vehicle Model to a Database (Using a Single Action).

To add a vehicle model to a database (using separate steps)

  1. Add a new vehicle model using the NewVehicleModel action. For example:

    curl http://localhost:14000/action=NewVehicleModel -F database=Cars -F identifier=FordFocus -F vehiclemake=Ford

    Media Server adds the vehicle model to the database and returns the identifier.

  2. Add one or more training images to the model using the AddVehicleModelImages action.

    For example, to add a training image by supplying the image data:

    curl http://localhost:14000/action=AddVehicleModelImages -F database=Cars -F identifier=FordFocus -F imagedata=@ford-focus.png -F anprlocations=DE

    Alternatively, to add a training image by supplying its path:

    curl http://localhost:14000/action=AddVehicleModelImages -F database=Cars -F identifier=FordFocus -F imagepath=./images/ford-focus.png -F anprlocations=DE
  3. (Optional) Add metadata to the vehicle model using the AddVehicleModelMetadata action. You can add any number of key-value pairs. For example:

    curl http://localhost:14000/action=AddVehicleModelMetadata -F database=Cars -F identifier=FordFocus -F key=type -F value=hatchback
  4. Complete the training for the vehicle model using the BuildVehicleModel action.

    curl http://localhost:14000/action=BuildVehicleModel -F database=Cars -F identifier=FordFocus