Annotating an image

AnnoStation is designed to foster the maximum productivity of annotators.

Steps

Here are some typical user steps to annotate an image in AnnoStation.

1. Open an image.

2. Select a label.

3. Select an appropriate tool.

4. Change the zoom level and annotate an object.

5. (Optional) provide attribute values.

6. Submit an image.

7. Move to the next image.

You can use keyboard shortcut keys to switch between these operations.

You can learn all supported keyboard shortcut keys here:

Keyboard shortcuts

Select a label

Select a label to start annotating an object from a list of labels defined for the Project.

Select a tool

You can use various tools, such as Path, Polygon, Bounding Box, etc to annotate an object. Available tools are different depending on if the selected label is segmentation or vector.

To learn more about supported tools, read here:

Annotation tools

Control the view of an image

To achieve precise annotation, you can adjust the zoom level of an image:

• Zoom in/out

• Pan

• Fit to the window

You can also use the Ctrl or ⌘ + Mouse Wheel to zoom in our out.

Provide attribute values

If attributes are associated with a vector label, you are asked to provide values as soon as you complete drawing a vector shape.

Submit an image

Once you are done with annotating all objects, you can submit an image. Once an image is submitted, you can no longer make modifications unless you are Developers or Managers so make sure you annotated everything.

You can also save an image before submitting an image.

To learn more about the workflow function, read here:

Working with a Project

Comment

Leaving a comment is a great way to raise a flag and avoid possible iterations in the later phase.

You can learn more about our comment features here:

Comments