Wavelet Toolbox™ provides apps and functions for analyzing and synthesizing signals and images. You can detect events like anomalies, change points, and transients, and denoise and compress data. Wavelet and other multiscale techniques can be used to analyze data at different time and frequency resolutions and to decompose signals and images into their various components. You can use wavelet techniques to reduce dimensionality and extract discriminating features from signals and images to train machine and deep learning models.
With Wavelet Toolbox you can interactively denoise signals, perform multiresolution and wavelet analysis, and generate MATLAB® code. The toolbox includes algorithms for continuous and discrete wavelet analysis, wavelet packet analysis, multiresolution analysis, wavelet scattering, and other multiscale analysis.
Many toolbox functions support C/C++ and CUDA® code generation for desktop prototyping and embedded system deployment.
Learn the basics of Wavelet Toolbox
CWT, constant-Q transform, empirical mode decomposition, wavelet coherence, wavelet cross-spectrum
DWT, MODWT, dual-tree wavelet transform, shearlets, wavelet packets, multisignal analysis
Wavelet shrinkage, nonparametric regression, block thresholding, multisignal thresholding
Wavelet-based techniques for machine learning and deep learning, GPU acceleration, hardware deployment, signal labeling
Orthogonal and biorthogonal wavelet and scaling filters, lifting
Generate C/C++ and CUDA code and MEX functions, and run functions on a graphics processing unit (GPU)