Z-Score Normalization
Standardizes the working data using MATLAB normalize with its default The transformation centers and scales values according to MATLAB's default operating dimension for the input array.
Overview
Standardizes the working data using MATLAB normalize with its default The transformation centers and scales values according to MATLAB's default operating dimension for the input array.
Current Idcubepro Implementation
The working cube is passed directly to normalize(Image).
- Standardized numerical values are useful for downstream analysis.
- Differences in absolute intensity scale should be reduced.
- Preparing data for some statistical or machine-learning workflows.
Important
Because a hyperspectral cube is multidimensional, the operating dimension used by normalize should be considered when interpreting the result.
Z-score normalization changes both offset and scale and therefore does not preserve the original physical intensity units.
Workflow
1. Load a hyperspectral dataset.
2. Select Z-score Normalization.
3. Allow processing to complete.
4. Inspect the normalized spectra and downstream results.
5. Use Reset Filters if original values are required.
Related Topics
- Standard Deviation Normalization