OpenCV checkChessboard_1
CheckChessboard_1 — the twin boolean, still just asking 'is it a chessboard?'
- img
- bool
checkChessboard_1 is a byte-for-byte duplicate of checkChessboard_0 - same img and size inputs, same single boolean output, same behavior. This is the pack's auto-generated twin pattern again: one node per type-definition overload, with the MatLike and UMat variants landing as identical _0/_1 pairs. There is no feature difference to find here, and you shouldn't spend another second looking for one.
What it does, for anyone who landed here first: it's a cheap test that answers "does this image contain a chessboard pattern of this grid size?" You give it img (NPARRAY) and size (STRING, as a Python literal like (9, 6) - parsed with literal_eval, so "9x6" will throw the pack's standard invalid syntax (unknown, line 0)). It returns bool. Its real use is camera calibration: filter a stack of photos, keep only the ones containing the pattern, then run the heavier corner-finding (findChessboardCorners) on those. In a pure image-generation workflow you will never open it.
Two things worth knowing that aren't obvious from the node itself:
- It's an
ximgprocfunction. This comes from OpenCV's extended image-processing module, which only ships in the contrib build - the exact reason this pack's install step ispip install opencv-python-contriband not plainopencv-python. If you only have the standard build, importing the module this node depends on fails at load. - The dependency conflict is real. The README's most-cited failure is
Cannot import name 'guidedFilter' from 'cv2.ximgproc'- same contrib module, same cause: two conflicting OpenCV wheels in one environment. The fix is to ensure exactly one OpenCV install, the contrib one, and restart ComfyUI. It's the most common broken-install symptom in the whole pack, and it usually bites people who installed multiple packs that each pulled a different opencv wheel.
Install is the standard pack story - ComfyUI Manager, search "OpenCV", or:
cd ComfyUI/custom_nodes
git clone https://github.com/geroldmeisinger/opencv-comfyui
pip install opencv-python-contrib
Restart, done, no model downloads. Remember the representation rules: NPARRAY in (BGR 0..255 uint8, batch size 1), bridge with Image2Nparray. And take the author's own advice to heart: auto-generated, ugly, "expect dragons." The duplicate-chessboard node is a gentle dragon.
Inputs (2)
| Name | Type | Default | Description |
|---|---|---|---|
| img | NPARRAY | — | |
| size | STRING | — |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| bool | BOOLEAN | — |