ComfyUI Node
CV Skeletonize
Reduces the foreground of a binary image to its 1-pixel-wide skeleton (centerline). Thinning (cv2.ximgproc, opencv-contrib) gives clean connected centerlines; 'morphological' is a coarser fallback that works without the contrib build. Used by the clock-reading pipeline to turn thick hands into measurable rays.
CV Skeletonize
- image
- skeleton
◄methodthinning (Zhang-Suen)►
Categoryimage/CV/contours
Inputs (2)
| Name | Type | Default | Description |
|---|---|---|---|
| image | COMFY_MATCHTYPE_V3 | Binary image (or MASK); non-zero = foreground. The skeleton comes back in this input's format. Accepts a ComfyUI IMAGE/MASK directly (frame 0 of a batch) or an NPARRAY. Arithmetic ops (add, multiply, etc.) process the full IMAGE batch when both inputs have the same batch size. | |
| method | COMBO | thinning (Zhang-Suen) | Zhang-Suen / Guo-Hall need opencv-contrib-python(-headless) (cv2.ximgproc) and give the cleanest centerlines; if that build is unavailable they fall back to 'morphological' automatically. Morphological works everywhere but is noisier. |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| skeleton | COMFY_MATCHTYPE_V3 | Same format as the image input. |