Normalization

Bases: ImageOnlyTransform

init ¶

__init__(mean: ndarray, std: ndarray, p: float = 1.0, copy: bool = False, **kwargs) -> None

Image level normalization transformation.

NOTE: this is not dataset-level normalization but image-level.

Parameters:

Name	Type	Description	Default
`mean`	`ndarray`	Mean values for each channel. Shape (C,)	required
`std`	`ndarray`	Standard deviation values for each channel. Shape (C,)	required
`p`	`float, default=1.0`	Probability of applying the transformation.	`1.0`
`copy`	`bool, default=False`	If True, normalize the copy of the input.	`False`

apply(image: ndarray, **kwargs) -> np.ndarray

Apply image-level normalization to input image.

Parameters:

Name	Type	Description	Default
`image`	`ndarray`	Input image to be normalized. Shape (H, W, C)\|(H, W).	required

Returns:

Type	Description
`ndarray`	np.ndarray: Normalized image. Same shape as input. dtype: float32.

get_transform_init_args_names()

Get the names of the transformation arguments.