Question 1

What is the best camera calibration pattern?

Accepted Answer

For most cameras a checkerboard or ChArUco board is the best starting point — they give accurate corners and are easy to detect. Use ChArUco when the board may be partly out of frame, ArUco grids for multi-marker setups, and circle grids when you want sub-pixel centroids in slightly blurry images.

Question 2

What's the difference between a checkerboard and a ChArUco board?

Accepted Answer

A checkerboard needs the full board visible to detect all corners. A ChArUco board adds unique ArUco markers inside the white squares, so it can still be identified and used for calibration even when partially occluded or cut off at the edges.

Question 3

ChArUco vs ArUco — which should I use?

Accepted Answer

A ChArUco board is a checkerboard fused with ArUco markers and gives the most accurate corners. A plain ArUco grid is a set of standalone markers — handy for partial views and pose estimation, but a ChArUco board is usually more accurate for intrinsic camera calibration.

Question 4

What size should calibration pattern squares be?

Accepted Answer

Make the board large enough to fill a good part of the frame at your working distance — typically 15–35 mm squares printed on A4/Letter or larger. The generator shows the exact board size in millimetres; enter that real-world size in CalibrX so the scale is correct.

Question 5

How do I print a calibration pattern correctly?

Accepted Answer

Print at 100% / "actual size" (disable "fit to page" or "scale to fit"), then measure the 50 mm scale bar with a ruler to confirm it matches. Mount the print flat on a rigid, non-glossy surface — any bend or ripple adds calibration error.

Question 6

Are these patterns compatible with OpenCV?

Accepted Answer

Yes. The generator uses the real OpenCV ArUco dictionaries, so every checkerboard, ChArUco, ArUco, and circle-grid target detects in OpenCV — or upload your photos of the printed board to CalibrX and calibrate online without writing any code.

Pattern	Best for	Corner accuracy	Partial / occluded views	Notes
Checkerboard	Simple, controlled setups	Highest	Must be fully visible	The classic target; every inner corner must be seen in each frame.
ChArUco	Most calibrations (recommended)	High	Robust	Checkerboard corners plus ArUco IDs — keeps sub-pixel accuracy even when partly out of frame.
ArUco	Cluttered or partial views	Medium	Robust	Each marker is identified individually, so occlusion is tolerated.
Circles grid	Industrial / machine vision	High (centroids)	Must be fully visible	Symmetric or asymmetric; sub-pixel centers but sensitive to strong perspective.

Camera calibration pattern generator

Which calibration pattern should you use?

Calibration patterns — common questions