public class ResidualsTriangulateSampson
extends java.lang.Object
implements org.ddogleg.optimization.functions.FunctionNtoM
Sampson first-order to geometric triangulation error. Partially enforces epipolar constraints. Much less expensive to compute than true constrained geometric error.
[1] Page 315 in R. Hartley, and A. Zisserman, "Multiple View Geometry in Computer Vision", 2nd Ed, Cambridge 2003
| Constructor and Description |
|---|
ResidualsTriangulateSampson() |
| Modifier and Type | Method and Description |
|---|---|
int |
getNumOfInputsN() |
int |
getNumOfOutputsM() |
void |
process(double[] input,
double[] output) |
void |
setObservations(java.util.List<georegression.struct.point.Point2D_F64> observations,
java.util.List<org.ejml.data.DenseMatrix64F> essential)
Configures inputs.
|
public void setObservations(java.util.List<georegression.struct.point.Point2D_F64> observations,
java.util.List<org.ejml.data.DenseMatrix64F> essential)
observations - Observations of the feature at different locations. Normalized image coordinates.essential - Essential matrices associated with camera motion (world to camera)public int getNumOfInputsN()
getNumOfInputsN in interface org.ddogleg.optimization.functions.FunctionNtoMpublic int getNumOfOutputsM()
getNumOfOutputsM in interface org.ddogleg.optimization.functions.FunctionNtoMpublic void process(double[] input,
double[] output)
process in interface org.ddogleg.optimization.functions.FunctionNtoM