Spectral Mean Filter
Smooths each pixel spectrum along the spectral dimension by averaging neighboring spectral bands. Spatial image geometry is not changed.
Overview
Smooths each pixel spectrum along the spectral dimension by averaging neighboring spectral bands. Spatial image geometry is not changed.
Filter Size defines the number of neighboring spectral bands included in the moving-average window.
If an even value is entered, IDCubePro increases it to the next Spectral averaging reduces rapid band-to-band fluctuations and can improve the apparent signal-to-noise ratio of spectra.
- Spectra contain high-frequency band-to-band noise.
- Mild spectral smoothing is required.
- Spatial image resolution should remain unchanged.
Large averaging windows can broaden spectral peaks, reduce narrow absorption features, and decrease effective spectral resolution.
Workflow
1. Load a hyperspectral dataset.
2. Open Filtering & Enhancement.
3. Select Spectral Mean Filter.
4. Set the desired spectral window size.
5. Compare the filtered and original spectral features.
6. Use Reset Filters if the result is not desired.
Related Topics
- First and Second Derivative Spectra