OpenCV findChessboardCornersSBWithMeta_1

findChessboardCornersSBWithMeta_1 is the second auto-generated overload of cv2.findChessboardCornersSBWithMeta, and it is, predictably by now, identical to findChessboardCornersSBWithMeta_0: same inputs (image, patternSize, flags, optional corners and meta), same three outputs (bool, corner nparray, quality-score nparray), same function. The overload numbering produced twins; there is no behavioral difference between them. Use whichever you grabbed.

The knowledge lives on the findChessboardCornersSBWithMeta_0 page. The useful summary: it's the robust saddle-point checkerboard detector (the same algorithm as findChessboardCornersSB) with an extra per-corner quality score in output nparray_2 - the confidence value that lets you filter or down-weight shaky corners before feeding them to a calibration solver. Remember patternSize is the interior-corner count as a literal tuple ((9, 6) for a 10×7 board, parentheses required), and that nparray_2 is a 1-D array of floats, not an image - don't route it to Nparrays2Image expecting a picture.

Installing it

Part of geroldmeisinger/opencv-comfyui - one install gets all ~600 nodes. ComfyUI Manager → search opencv-comfyui, or:

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

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

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

That's it - a faithful twin. When you need checkerboard corners with confidence scores, use the _0 page's guidance and either of the twins.