Geometric (Box, Polygon, Line, Keypoints, Cuboid) Annotations

In this guide, we'll use the Python SDK to construct geometric annotations. Each geometric
annotation object in Nucleus has five components:

Label of the annotation in the class taxonomy, e.g. car or pedestrian
Geometry to define the position, size, etc. of the box / polygon / line / cuboid
Reference ID of the item to which to apply the annotation
Annotation ID, an optional unique (per dataset item) key to refer to the annotation
Metadata, optional key-value pairs pertaining to the annotation, e.g. vehicle_color = blue

All geometric annotation types (box, polygon, line, cuboid) share the same structure aside from the geometry. Below, we'll explore how to construct the annotation objects, including geometries, for each type:

`BoxAnnotation`

A 2D bounding box based on a given XY position, width, and height.

For more info, see our Python SDK Reference for BoxAnnotations.

from nucleus import BoxAnnotation

box_gt = BoxAnnotation(
    label="car",
    x=4,        # px between left edge of the box and left border of the image
    y=6,        # px between top edge of the box and top border of the image
    width=12,   # width in px
    height=18,  # height in px
    reference_id="image_1",
    annotation_id="image_1_car_box_1",
    metadata={"vehicle_color": "ruby"}
)

`PolygonAnnotation`

A 2D polygon constructed by connecting 3+ vertices.

For more info, see our Python SDK Reference for PolygonAnnotations.

from nucleus import Point, PolygonAnnotation

polygon_gt = PolygonAnnotation(
    label="bus",
    vertices=[	# list of vertices as Point objects
        Point(x=100, y=100),
        Point(150, 200),
        Point(200, 100),
    ],
    reference_id="image_2",
    annotation_id="image_2_bus_polygon_1",
    metadata={"vehicle_color": "sapphire"}
)

`LineAnnotation`

A 2D line constructed by connecting 2+ vertices (an unclosed polygon).

For more info, see our Python SDK Reference for LineAnnotations.

from nucleus import Point, LineAnnotation

line_gt = LineAnnotation(
    label="lane",
    vertices=[	# list of vertices as Point objects
        Point(x=100, y=100),
        Point(200, 300),
        Point(300, 200),
    ],
    reference_id="image_1",
    annotation_id="image_1_lane_line_1",
    metadata={"position": "shoulder"},
)

`KeypointsAnnotation`

A collection of 2D points that are named and connected via a domain-specific skeleton.

For more info, see our Python SDK Reference for KeypointsAnnotations.

from nucleus import Keypoint, KeypointsAnnotation

keypoint_gt = KeypointsAnnotation(
    label="face",
    keypoints=[
        Keypoint(100, 100, visible=True),
        Keypoint(200, 300, visible=False),
        Keypoint(visible=False),
    ],
    names=["left_eye", "right_eye", "nose"],
    skeleton=[[0, 2], [1, 2]],
    reference_id="image_1",
    annotation_id="image_1_face_1",
    metadata={"visibility": "partial"},
)

`CuboidAnnotation`

A 3D cuboid based on a given position, dimensions, and yaw. Cannot be applied to 2D images, only 3D pointcloud DatasetItems.

For more info, see our Python SDK Reference for CuboidAnnotations.

from nucleus import Point3D, CuboidAnnotation

cuboid_gt = CuboidAnnotation(
    label="car",
    position=Point3D(100, 100, 10),			# center of the cuboid
    dimensions=Point3D(x=5, y=10, z=5),	# length (x), width (y), height (z)
    yaw=0,															# rotation in rad about the Z axis
    reference_id="pointcloud_1",
    annotation_id="pointcloud_1_car_cuboid_1",
    metadata={"vehicle_color": "emerald"}
)