unitxt.image_operators module

class unitxt.image_operators.DecodeImage(data_classification_policy: List[str] = None, _requirements_list: List[str] | Dict[str, str] = [], requirements: List[str] | Dict[str, str] = [], caching: bool = None, apply_to_streams: List[str] = None, dont_apply_to_streams: List[str] = None, field: str | NoneType = None, to_field: str | NoneType = None, field_to_field: List[List[str]] | Dict[str, str] | NoneType = None, use_query: bool | NoneType = None, process_every_value: bool = False, get_default: Any = None, not_exist_ok: bool = False, not_exist_do_nothing: bool = False)

Bases: FieldOperator, PillowMixin

class unitxt.image_operators.EncodeImageToString(data_classification_policy: List[str] = None, _requirements_list: List[str] | Dict[str, str] = [], requirements: List[str] | Dict[str, str] = [], caching: bool = None, apply_to_streams: List[str] = None, dont_apply_to_streams: List[str] = None, field: str | NoneType = None, to_field: str | NoneType = None, field_to_field: List[List[str]] | Dict[str, str] | NoneType = None, use_query: bool | NoneType = None, process_every_value: bool = False, get_default: Any = None, not_exist_ok: bool = False, not_exist_do_nothing: bool = False, image_format: str = 'JPEG')

Bases: FieldOperator

class unitxt.image_operators.GrayScale(data_classification_policy: List[str] = None, _requirements_list: Union[List[str], Dict[str, str]] = [], requirements: Union[List[str], Dict[str, str]] = [], caching: bool = None, apply_to_streams: List[str] = None, dont_apply_to_streams: List[str] = None, field: Union[str, NoneType] = None, to_field: Union[str, NoneType] = None, field_to_field: Union[List[List[str]], Dict[str, str], NoneType] = None, use_query: Union[bool, NoneType] = None, process_every_value: bool = False, get_default: Any = None, not_exist_ok: bool = False, not_exist_do_nothing: bool = False, augmented_type: object = <class 'unitxt.types.Image'>)

Bases: ImageAugmentor

class unitxt.image_operators.GridLines(data_classification_policy: List[str] = None, _requirements_list: Union[List[str], Dict[str, str]] = [], requirements: Union[List[str], Dict[str, str]] = [], caching: bool = None, apply_to_streams: List[str] = None, dont_apply_to_streams: List[str] = None, field: Union[str, NoneType] = None, to_field: Union[str, NoneType] = None, field_to_field: Union[List[List[str]], Dict[str, str], NoneType] = None, use_query: Union[bool, NoneType] = None, process_every_value: bool = False, get_default: Any = None, not_exist_ok: bool = False, not_exist_do_nothing: bool = False, augmented_type: object = <class 'unitxt.types.Image'>, num_lines: int = 128, line_thickness: int = 1, line_color: Tuple[int, int, int] = (255, 255, 255))

Bases: ImageAugmentor

A class that overlays a fixed number of evenly spaced horizontal and vertical lines on an image.

Parameters:

• num_lines (int) – The number of horizontal and vertical lines to add.

• line_thickness (int) – Thickness of each line in pixels.

• line_color (Tuple[int, int, int]) – RGB color of the grid lines.

process_image(image)

Adds grid lines to the provided image and returns the modified image.

line_color: Tuple[int, int, int] = (255, 255, 255)

class unitxt.image_operators.ImageAugmentor(data_classification_policy: List[str] = None, _requirements_list: Union[List[str], Dict[str, str]] = [], requirements: Union[List[str], Dict[str, str]] = [], caching: bool = None, apply_to_streams: List[str] = None, dont_apply_to_streams: List[str] = None, field: Union[str, NoneType] = None, to_field: Union[str, NoneType] = None, field_to_field: Union[List[List[str]], Dict[str, str], NoneType] = None, use_query: Union[bool, NoneType] = None, process_every_value: bool = False, get_default: Any = None, not_exist_ok: bool = False, not_exist_do_nothing: bool = False, augmented_type: object = <class 'unitxt.types.Image'>)

Bases: TaskInputsAugmentor, PillowMixin

augmented_type

alias of Image

class unitxt.image_operators.ImageDataString

Bases: str

class unitxt.image_operators.ImageFieldOperator(data_classification_policy: List[str] = None, _requirements_list: List[str] | Dict[str, str] = [], requirements: List[str] | Dict[str, str] = [], caching: bool = None, apply_to_streams: List[str] = None, dont_apply_to_streams: List[str] = None, field: str | NoneType = None, to_field: str | NoneType = None, field_to_field: List[List[str]] | Dict[str, str] | NoneType = None, use_query: bool | NoneType = None, process_every_value: bool = False, get_default: Any = None, not_exist_ok: bool = False, not_exist_do_nothing: bool = False)

Bases: FieldOperator, PillowMixin

class unitxt.image_operators.Oldify(data_classification_policy: List[str] = None, _requirements_list: Union[List[str], Dict[str, str]] = [], requirements: Union[List[str], Dict[str, str]] = [], caching: bool = None, apply_to_streams: List[str] = None, dont_apply_to_streams: List[str] = None, field: Union[str, NoneType] = None, to_field: Union[str, NoneType] = None, field_to_field: Union[List[List[str]], Dict[str, str], NoneType] = None, use_query: Union[bool, NoneType] = None, process_every_value: bool = False, get_default: Any = None, not_exist_ok: bool = False, not_exist_do_nothing: bool = False, augmented_type: object = <class 'unitxt.types.Image'>, noise_strength: int = 30, tint_strength: float = 0.4)

Bases: ImageAugmentor

class unitxt.image_operators.PillowMixin(data_classification_policy: List[str] = None, _requirements_list: List[str] | Dict[str, str] = {'PIL': 'pip install pillow'}, requirements: List[str] | Dict[str, str] = [])

Bases: PackageRequirementsMixin

class unitxt.image_operators.PixelNoise(data_classification_policy: List[str] = None, _requirements_list: Union[List[str], Dict[str, str]] = [], requirements: Union[List[str], Dict[str, str]] = [], caching: bool = None, apply_to_streams: List[str] = None, dont_apply_to_streams: List[str] = None, field: Union[str, NoneType] = None, to_field: Union[str, NoneType] = None, field_to_field: Union[List[List[str]], Dict[str, str], NoneType] = None, use_query: Union[bool, NoneType] = None, process_every_value: bool = False, get_default: Any = None, not_exist_ok: bool = False, not_exist_do_nothing: bool = False, augmented_type: object = <class 'unitxt.types.Image'>, square_size: int = 1, noise_rate: float = 0.3)

Bases: ImageAugmentor

A class that overlays a mask of randomly colored nxn squares across an image based on a specified noise rate.

Parameters:

• square_size (int) – Size of each square in pixels.

• noise_rate (float) – Proportion of the image that should be affected by noise (0 to 1).

process_image(image)

Adds the random square mask to the provided image and returns the modified image.

class unitxt.image_operators.ToImage(data_classification_policy: List[str] = None, _requirements_list: List[str] | Dict[str, str] = [], requirements: List[str] | Dict[str, str] = [], caching: bool = None, apply_to_streams: List[str] = None, dont_apply_to_streams: List[str] = None, field: str | NoneType = None, to_field: str | NoneType = None, field_to_field: List[List[str]] | Dict[str, str] | NoneType = None, use_query: bool | NoneType = None, process_every_value: bool = False, get_default: Any = None, not_exist_ok: bool = False, not_exist_do_nothing: bool = False)

Bases: InstanceFieldOperator

class unitxt.image_operators.ToRGB(data_classification_policy: List[str] = None, _requirements_list: List[str] | Dict[str, str] = [], requirements: List[str] | Dict[str, str] = [], caching: bool = None, apply_to_streams: List[str] = None, dont_apply_to_streams: List[str] = None, field: str | NoneType = None, to_field: str | NoneType = None, field_to_field: List[List[str]] | Dict[str, str] | NoneType = None, use_query: bool | NoneType = None, process_every_value: bool = False, get_default: Any = None, not_exist_ok: bool = False, not_exist_do_nothing: bool = False)

Bases: ImageFieldOperator

unitxt.image_operators.data_url_to_image(data_url: str)

unitxt.image_operators.extract_images(instance)

unitxt.image_operators.image_to_data_url(image: Image, default_format='JPEG')

Convert an image to a data URL.

https://developer.mozilla.org/en-US/docs/Web/URI/Schemes/data