Class DistanceHomographySq

All Implemented Interfaces:

public class DistanceHomographySq extends Object implements DistanceFromModel<Homography2D_F64,AssociatedPair>

Computes the Euclidean error squared between 'p1' and 'p2' after projecting 'p1' into image 2. Input can be in pixels or normalized image coordinates, but the error for normalized image coordinates doesn't have a physical meaning.

error = (p2'.x - p2.x)2 + (p2'.y - p2.y)2, where p2' is the predicted location and p2 is the observed location.