RandomCropPerturber

class nrtk.impls.perturb_image.geometric.random.RandomCropPerturber(*, crop_size: tuple[int, int] | None = None, seed: int | None = None, is_static: bool = False)

RandomCropPerturber randomly crops an image and adjusts bounding boxes accordingly.

Attributes:

crop_size (tuple[int, int]): Target crop dimensions for the input image. seed (int | None): Random seed for reproducibility. None for non-deterministic behavior. is_static (bool): If True, resets RNG after each call for consistent results.

Methods

from_config	Instantiate a new instance of this class given the configuration JSON-compliant dictionary encapsulating initialization arguments.
get_config	Returns the current configuration of the RandomCropPerturber instance.
get_default_config	Generate and return a default configuration dictionary for this class.
get_impls	Discover plugins, skipping any entrypoints that fail to load.
get_type_string	Returns the fully qualified type string of the PerturbImage class or its subclass.
is_usable	Check whether this class is available for use.
perturb	Randomly crops an image and adjusts bounding boxes.

__init__(*, crop_size: tuple[int, int] | None = None, seed: int | None = None, is_static: bool = False) → None

RandomCropPerturber applies a random cropping perturbation to an input image.

It ensures that bounding boxes are adjusted correctly to reflect the new cropped region.

Args:

crop_size:

Target crop size as (crop_height, crop_width). If crop_size is None, it defaults to the size of the input image. If crop_size is greater than or equal to image size, the original image is returned.

seed:

Random seed for reproducible results. Defaults to None for non-deterministic behavior.

is_static:

If True and seed is provided, resets RNG after each perturb call for consistent results across multiple calls (useful for video frame processing).

get_config() → dict[str, Any]

Returns the current configuration of the RandomCropPerturber instance.

perturb(*, image: ndarray[Any, Any], boxes: Iterable[tuple[AxisAlignedBoundingBox, dict[Hashable, float]]] | None = None, **_: Any) → tuple[ndarray[Any, Any], Iterable[tuple[AxisAlignedBoundingBox, dict[Hashable, float]]] | None]

Randomly crops an image and adjusts bounding boxes.

Args:

image:

Input image as a numpy array of shape (H, W, C).

boxes:

List of bounding boxes in AxisAlignedBoundingBox format and their corresponding classes.

Returns:

Cropped image with the modified bounding boxes.