nrtk.impls.perturb_image.generic.skimage.random_noise

This module provides a set of classes for adding different types of noise to images, implementing the PerturbImage interface. The perturbation types include salt, pepper, salt-and-pepper, Gaussian, and speckle noise, allowing for a wide range of image noise simulations.

Classes:

_SKImageNoisePerturber: Base class for noise perturbation, using skimage.util.random_noise. _SPNoisePerturber: Base class for salt-and-pepper type noise. SaltNoisePerturber: Adds salt noise to an image. PepperNoisePerturber: Adds pepper noise to an image. SaltAndPepperNoisePerturber: Adds both salt and pepper noise to an image, with control over

the ratio of salt to pepper.

_GSNoisePerturber: Base class for Gaussian-based noise. GaussianNoisePerturber: Adds Gaussian-distributed additive noise to an image. SpeckleNoisePerturber: Adds Gaussian-based multiplicative noise (speckle) to an image.

Dependencies:

• numpy for handling image data and random number generation.

• skimage.util for applying various noise effects to images.

Usage:

Each noise perturber class can be instantiated with specific parameters, allowing the user to customize the type and intensity of noise applied to an image. Use the perturb method of each class to apply the chosen noise effect.

Example:

gaussian_perturber = GaussianNoisePerturber(mean=0, var=0.01) noisy_image, boxes = gaussian_perturber.perturb(image_data, boxes)

Notes:

• The boxes returned from perturb are identical to the boxes passed in.

Classes

GaussianNoisePerturber	Adds Gaussian-distributed additive noise to image stimulus.
PepperNoisePerturber	Adds pepper noise to image stimulus.
SaltAndPepperNoisePerturber	Adds salt & pepper noise to image stimulus.
SaltNoisePerturber	Adds salt noise to image stimulus.
SpeckleNoisePerturber	Adds multiplicative noise to image stimulus.