SlidingWindowGeoDataset#

class SlidingWindowGeoDataset[source]#

Bases: GeoDataset

Read the scene left-to-right, top-to-bottom, using a sliding window.

__init__(scene: Scene, size: Union[int, tuple[int, int]], stride: Union[int, tuple[int, int]], out_size: Optional[Union[int, tuple[int, int]]] = None, padding: Optional[Union[int, tuple[int, int]]] = None, pad_direction: Literal['both', 'start', 'end'] = 'end', within_aoi: bool = True, transform: albumentations.core.transforms_interface.BasicTransform | None = None, transform_type: rastervision.pytorch_learner.dataset.transform.TransformType | None = None, normalize: bool = True, to_pytorch: bool = True, return_window: bool = False)[source]#

Constructor.

Parameters:

  • object. (scene A Scene) –

  • size (Union[int, tuple[int, int]]) – Window size.

  • stride (Union[int, tuple[int, int]]) – Step size between windows.

  • to (out_size Resize chips to this size before returning. Defaults) – None.

  • padding (Optional[Union[int, tuple[int, int]]]) – How many pixels the windows are allowed to overflow the sides of the raster source. If None, will be automatically calculated such that the windows cover the entire extent. Defaults to None.

  • pad_direction (Literal['both', 'start', 'end']) – If 'end', only pad ymax and xmax (bottom and right). If 'start', only pad ymin and xmin (top and left). If 'both', pad all sides. If 'both' pad all sides. Has no effect if padding is zero. Defaults to 'end'.

  • AOI (within_aoi If True and if the scene has an) – windows that lie fully within the AOI. If False, windows only partially intersecting the AOI will also be allowed. Defaults to True.

  • sample (only) – windows that lie fully within the AOI. If False, windows only partially intersecting the AOI will also be allowed. Defaults to True.

  • transform (albumentations.core.transforms_interface.BasicTransform | None) – Albumentations transform to apply to the windows. Defaults to None. Each transform in Albumentations takes images of type uint8, and sometimes other data types. The data type requirements can be seen at https://albumentations.ai/docs/api_reference/augmentations/transforms/ # noqa If there is a mismatch between the data type of imagery and the transform requirements, a RasterTransformer should be set on the RasterSource that converts to uint8, such as MinMaxTransformer or StatsTransformer.

  • transform_type (rastervision.pytorch_learner.dataset.transform.TransformType | None) – Type of transform. Defaults to None.

  • normalize (bool) – If True, the sampled chips are normalized to [0, 1] based on their data type. Defaults to True.

  • to_pytorch (bool) – If True, the sampled chips and labels are converted to pytorch tensors. Defaults to True.

  • return_window (bool) – Make __getitem__ return the window coordinates used to generate the image. Defaults to False.

  • scene (Scene) –

  • out_size (Optional[Union[int, tuple[int, int]]]) –

  • within_aoi (bool) –

Methods

__init__(scene, size, stride[, out_size, ...])

Constructor.

append_resize_transform(transform, out_size)

Get transform to use for resizing windows to out_size.

from_uris(*args, **kwargs)

init_windows()

Pre-compute windows.
__add__(other: Dataset[T_co]) → ConcatDataset[T_co]#
Parameters:

other (Dataset[T_co]) –

Return type:

ConcatDataset[T_co]

__getitem__(idx: int)[source]#
Parameters:

idx (int) –

__init__(scene: Scene, size: Union[int, tuple[int, int]], stride: Union[int, tuple[int, int]], out_size: Optional[Union[int, tuple[int, int]]] = None, padding: Optional[Union[int, tuple[int, int]]] = None, pad_direction: Literal['both', 'start', 'end'] = 'end', within_aoi: bool = True, transform: albumentations.core.transforms_interface.BasicTransform | None = None, transform_type: rastervision.pytorch_learner.dataset.transform.TransformType | None = None, normalize: bool = True, to_pytorch: bool = True, return_window: bool = False)[source]#

Constructor.

Parameters:

  • object. (scene A Scene) –

  • size (Union[int, tuple[int, int]]) – Window size.

  • stride (Union[int, tuple[int, int]]) – Step size between windows.

  • to (out_size Resize chips to this size before returning. Defaults) – None.

  • padding (Optional[Union[int, tuple[int, int]]]) – How many pixels the windows are allowed to overflow the sides of the raster source. If None, will be automatically calculated such that the windows cover the entire extent. Defaults to None.

  • pad_direction (Literal['both', 'start', 'end']) – If 'end', only pad ymax and xmax (bottom and right). If 'start', only pad ymin and xmin (top and left). If 'both', pad all sides. If 'both' pad all sides. Has no effect if padding is zero. Defaults to 'end'.

  • AOI (within_aoi If True and if the scene has an) – windows that lie fully within the AOI. If False, windows only partially intersecting the AOI will also be allowed. Defaults to True.

  • sample (only) – windows that lie fully within the AOI. If False, windows only partially intersecting the AOI will also be allowed. Defaults to True.

  • transform (albumentations.core.transforms_interface.BasicTransform | None) – Albumentations transform to apply to the windows. Defaults to None. Each transform in Albumentations takes images of type uint8, and sometimes other data types. The data type requirements can be seen at https://albumentations.ai/docs/api_reference/augmentations/transforms/ # noqa If there is a mismatch between the data type of imagery and the transform requirements, a RasterTransformer should be set on the RasterSource that converts to uint8, such as MinMaxTransformer or StatsTransformer.

  • transform_type (rastervision.pytorch_learner.dataset.transform.TransformType | None) – Type of transform. Defaults to None.

  • normalize (bool) – If True, the sampled chips are normalized to [0, 1] based on their data type. Defaults to True.

  • to_pytorch (bool) – If True, the sampled chips and labels are converted to pytorch tensors. Defaults to True.

  • return_window (bool) – Make __getitem__ return the window coordinates used to generate the image. Defaults to False.

  • scene (Scene) –

  • out_size (Optional[Union[int, tuple[int, int]]]) –

  • within_aoi (bool) –

append_resize_transform(transform: albumentations.core.transforms_interface.BasicTransform | None, out_size: tuple[int, int]) → albumentations.augmentations.geometric.resize.Resize | albumentations.core.composition.Compose#

Get transform to use for resizing windows to out_size.

Parameters:

  • transform (albumentations.core.transforms_interface.BasicTransform | None) –

  • out_size (tuple[int, int]) –

Return type:

albumentations.augmentations.geometric.resize.Resize | albumentations.core.composition.Compose

classmethod from_uris(*args, **kwargs) → Self#
Return type:

Self

init_windows() → None[source]#

Pre-compute windows.

Return type:

None