whylogs.core.image_profiling

Module Contents

Classes

TrackImage

This is a class that computes image features and visits profiles and so image features can be sketched.

Functions

image_loader(path: str = None) → PIL.Image.Image

get_pil_image_metadata(img: PIL.Image.Image) → Dict

Grab metra data from a PIL Image

image_based_metadata(img)

Attributes

logger

ImageType

DEFAULT_IMAGE_FEATURES

_METADATA_DEFAULT_ATTRIBUTES

whylogs.core.image_profiling.logger
whylogs.core.image_profiling.ImageType
whylogs.core.image_profiling.DEFAULT_IMAGE_FEATURES
whylogs.core.image_profiling._METADATA_DEFAULT_ATTRIBUTES = ['ImageWidth', 'ImageLength', 'BitsPerSample', 'Compression', 'Quality',...
whylogs.core.image_profiling.image_loader(path: str = None)PIL.Image.Image
class whylogs.core.image_profiling.TrackImage(filepath: str = None, img: PIL.Image.Image = None, feature_transforms: List[Callable] = DEFAULT_IMAGE_FEATURES, feature_name: str = '', metadata_attributes: Union[str, List[str]] = _METADATA_DEFAULT_ATTRIBUTES)

This is a class that computes image features and visits profiles and so image features can be sketched.

feature_name

name given to this image feature, will prefix all image based features

Type

str

feature_transforms

Feature transforms to be apply to image data.

Type

List[Callable]

img

the PIL.Image

Type

PIL.Image

metadata_attributes

metadata attributes to track

Type

TYPE

__call__(self, profiles)

Call method to add image data and metadata to associated profiles :param profiles: DatasetProfile :type profiles: Union[List[DatasetProfile],DatasetProfile]

whylogs.core.image_profiling.get_pil_image_metadata(img: PIL.Image.Image)Dict

Grab metra data from a PIL Image

Parameters

img (ImageType) – PIL Image

Returns

of metadata

Return type

Dict

whylogs.core.image_profiling.image_based_metadata(img)