Our world is filled with data in the form of signals and images. This abundance of data makes it important to extract the essential information and disregard unimportant content when processing signals and images. In some cases, this means that you need to create sparse representations that eliminate all unnecessary detail. In other cases, you need to create redundant, or expansive, representations of the data so you can separate out the important features.
For example, you might need to:
- Remove noise while preserving important features
- Easily identify and extract discriminatory features
- Compress data while retaining the important information
- Determine the optimal representation of your data
Wavelet Toolbox provides apps and functions that enable you to easily analyze real-world signals and images. With the Wavelet Signal Denoiser App, you can automatically remove the unwanted components present in signals while preserving the sharp features. The toolbox also lets you:
- Analyze the variability present in signals at multiple scales
- Obtain discriminatory information from signals using fractal analysis, and
- Localize transients and changepoints in time-series data
The toolbox also enables you to:
- Explore how the frequency of a signal changes over time
- Filter time-localized frequency components in signals
- Reconstruct individual oscillatory modes from signals, and
- Identify coherent time-varying oscillations in two signals
Here, the estimates within the cone are reliable, and the arrows help determine the relative lag between the signals.
You can also analyze images using Wavelet Toolbox. For example, you can:
- Denoise images while preserving the edges
- Compress images while maintaining perceptual quality, and
- Analyze images in space, frequency, and orientation
The toolbox supports a number of wavelet families for performing wavelet analysis. For more information, return to the product page.
