OpenCV findChessboardCornersSB_1

By now you know the drill: findChessboardCornersSB_1 is the second auto-generated overload of cv2.findChessboardCornersSB, and it's identical to findChessboardCornersSB_0 in the graph - same inputs (image, patternSize, flags, optional corners), same outputs (bool + corner nparray), same saddle-point detection underneath. The pack generator numbered OpenCV's overloads, both schemas landed the same, and you get a pair of interchangeable nodes. Pick one.

The content that matters is on the findChessboardCornersSB_0 page, and it's worth reading if you're doing calibration: this is the robust checkerboard detector - it finds saddle points in the board's intensity landscape and validates their grid structure, which makes it dramatically more forgiving than the classic findChessboardCorners on occluded, glossy, or asymmetric boards. Inputs to know: patternSize as a literal interior-corner tuple like (9, 6) (squares count minus one per side, parentheses required), and flags - 0 for default, 8 for a fast pre-check. The bool output tells you whether the board was actually found.

Installing it

It ships with geroldmeisinger/opencv-comfyui, all ~600 nodes in one install. ComfyUI Manager → search opencv-comfyui, or:

cd ComfyUI/custom_nodes
git clone https://github.com/geroldmeisinger/opencv-comfyui

then restart. Needs opencv-contrib-python, numpy, torch. If startup throws Cannot import name 'guidedFilter' from 'cv2.ximgproc', conflicting OpenCV installs - consolidate to one.

House rules: nparrays in and out (Image2Nparray / Nparrays2Image), batch size 1 only.

That's the whole story. A faithful twin with nothing to add - use the _0 page's guidance and whichever node you grabbed.