N-33 Mask Edge Smoother
JHPixelProMaskEdgeSmoother cleans up noisy or jagged mask boundaries with bilateral filtering and an optional image guide. It is designed for the last stage of a masking pipeline, after person selection, morphology, or trimap work, when you want a more natural edge transition without rebuilding the mask from scratch.
Schema
| Name | Kind | Type / default | Description |
|---|---|---|---|
mask |
Input | MASK |
Source mask to smooth. |
guide |
Optional input | IMAGE |
Optional image guide for edge-aware smoothing when subject detail should steer the result. |
sigma_spatial |
Widget | FLOAT, default 4.0 |
Spatial radius of the bilateral smoothing pass. |
sigma_range |
Widget | FLOAT, default 0.1 |
Range-domain smoothing weight on the normalized mask values. |
iterations |
Widget | INT, default 1 |
Number of smoothing passes to run. |
mask |
Output | MASK |
Smoothed mask for downstream blending or compositing. |
Workflow preview
Workflow JSON: workflows/N-33-mask-edge-smoother.json