Working with Images

Wavelet image scattering, 2-D continuous wavelet transform, shearlets, stationary wavelet transform

Use shearletSystem to obtain directionally sensitive sparse representations of real-valued images with anisotropic features. Wavelet image scattering yields representations insensitive to image translations and rotations without sacrificing class discriminability.

Functions

Differentiable Signal Processing

`dldwt`	Differentiable discrete wavelet transform (Since R2025a)
`dlidwt`	Differentiable inverse discrete wavelet transform (Since R2025a)
`waveletPooling2dLayer`	2-D discrete wavelet pooling layer (Since R2026a)

Feature Extraction

`shearletSystem`	Cone-adapted bandlimited shearlet system
`waveletScattering2`	Wavelet image scattering

Datastores

`imageDatastore`	Datastore for image data
`augmentedImageDatastore`	Transform batches to augment image data

Augment Images

imageDataAugmenter Configure image data augmentation

Related Information

Wavelet Scattering

Featured Examples

Digit Classification with Wavelet Scattering

Digit Classification with Wavelet Scattering

Use wavelet scattering for image classification.

Open Live Script

Parasite Classification Using Wavelet Scattering and Deep Learning

Parasite Classification Using Wavelet Scattering and Deep Learning

Classify parasitic infections using wavelet image scattering and deep learning.

Since R2021b
Open Live Script

Texture Classification with Wavelet Image Scattering

Texture Classification with Wavelet Image Scattering

Extract wavelet-based features on a GPU for image classification.

Open Live Script