Using Vision Tools with Nonlinear Transforms

The client coordinate transform that is part of every pel buffer can be either linear (cc2XformLinear) or nonlinear (cc2XformPoly). All CVL vision tools support linear client coordinate transforms but only certain tools also support nonlinear client coordinate transforms.

The following CVL vision tools can make use of nonlinear transforms to return higher accuracy results than can be obtained using a linear transform:
CNLSearch
Blob
Caliper
PatMax

You can use the CVL Grid-of-Dots Calibration tool (described in the CVL Vision Tools Guide) to create a cc2XformPoly transformation (described in Math Foundations of Transformations). A polynomial nonlinear transformation uses a first, third, or fifth order polynomial equation to represent the slight nonlinear distortions introduced by many optical systems.

This section describes some general issues related to using nonlinear transformations with CVL vision tools, and it describes the specific approach taken by each of the tools that support the use of nonlinear transformations.