Checkerboard Calibration StepsCognex VisionPro

The Checkerboard Calibration tool operates, at both calibration time and run time, by executing a series of steps. Each of these steps consumes some input values and produces some output values. Use this topic if you want to understand the types of change events fired by the tool. Most VisionPro users who use the Checkerboard Calibration tool will not need to understand the tool to this detail.

See the following sections for more information:

Step Summary

The following 5 steps are performed during the calibration phase:

  1. Find and label the calibration plate grid points (tile vertices or dot centers) in the calibration image.
  2. Compute a (possibly nonlinear) calibration transform.
  3. If nonlinear, generate an undistorted calibration image.
  4. If nonlinear, select a destination rectangle and generate internal data structures for warping.
  5. Compose the linear adjustment transform.

Each step's inputs and outputs are listed in the following sections. The table shown below defines which outputs become invalid when a given step's inputs are changed.

Table 1. Step inputs
Changed Inputs to These Steps...Invalidate Outputs of These Steps
11, 2, 3, 4, 5
22, 3, 4, 5
33, 4, 5
44, 5
55
Step 1: Find and Label the Checkerboard Vertices

This step locates the vertices in the image of the checkerboard calibration plate or the dot centers in a grid-of-dots plate. It then uses the grid spacing ("tile size") information and the grid point locations to automatically label each grid point.

Inputs:

Outputs:

Step 2: Compute the Calibration Transform

This step computes the transformation between the uncalibrated and raw calibrated space.

Inputs:

Outputs:

Step 3: Generate the Undistorted Calibration Image

This step, which is only performed in nonlinear calibration, constructs internal data structures and warps the calibration image to produce the undistorted calibration image.

Inputs:

Outputs:

Step 4: Generate Data Structures for Run-Time Warping

This step, which is only performed in nonlinear calibration, constructs the internal data structures used to produce the warped (undistorted) run-time image.

Input:

Outputs:

Step 5: Compose the Adjustment Transform

This step composes the transformation computed in Step 2 with the user-supplied adjustment transformation.

Inputs:

Outputs: