SSTV Denoising
Applies spatial-spectral total-variation denoising to reduce noise while preserving important spatial and spectral structure.
Overview
Applies spatial-spectral total-variation denoising to reduce noise while preserving important spatial and spectral structure.
Lambda controls one of the regularization terms used by the SSTV Mu controls an additional regularization contribution in the SSTV Nu controls another regularization contribution used to balance spatial and spectral smoothing.
Iterations controls the optimization length. More iterations may improve convergence but increase processing time.
Light: weaker regularization and fewer iterations.
Balanced: general-purpose default.
Strong: stronger regularization and more iterations.
- Noise is present across spatial and spectral dimensions.
- Piecewise-smooth spatial structure should be preserved.
- Spectral noise reduction is required without simple averaging.
Excessive regularization can flatten subtle spatial textures, weak features, or meaningful spectral variation.
Workflow
1. Load a hyperspectral dataset.
2. Open Filtering & Enhancement.
3. Choose Light, Balanced, or Strong.
4. Adjust Lambda, Mu, Nu, or Iterations if required.
5. Run the denoising operation.
6. Evaluate image quality and spectral preservation.
7. Use Reset Filters if the result is not desired.