Checkerboard Targets

The checkerboard pattern consists of alternating black and white squares. Corner points are located at the black-and-white boundaries, and their sub-pixel coordinates are extracted using Harris or Shi-Tomasi corner detection. Corner extraction accuracy: 0.01~0.05 pixels (theoretically).

Each corner point of the checkerboard calibration board is a "precise physical coordinate reference point." Since the boundaries of the black and white squares are absolutely straight and precisely equidistant in the physical world (factory machining accuracy can reach ±0.001mm), the spatial positions of these corner points are "absolutely known." After the camera captures an image, these corner points are projected as pixels. By establishing the correspondence between "physical coordinates → pixel coordinates," the camera's intrinsic and extrinsic parameters can be deduced.

Key Specifications:

Size: The calibration board diagonally should cover 60%~80% of the field of view. Too small a size will reduce calibration accuracy; too large a size will not cover the entire field of view.

Cube Size: The cube should cover 20~50 pixels. Too large a size is detrimental to sub-pixel positioning; too small a size is easily affected by noise.

Accuracy Grade: Industrial-grade accuracy is ±0.025mm (corresponding to RMS < 0.3 pixels).

Applicable Scenarios:
General industrial inspection (preferred choice)
Area scan camera calibration
Monocular vision calibration
Laboratory precision measurement

Unsuitable Scenarios:
Scenarios with large tilt angles (>30°) (corner points may be occluded or distorted)
Low-light environments (insufficient corner contrast)
Scenarios with reflective or transparent objects (requires backlight transmissive type)