Public Member Functions | |
| ICPMethod (int value) | |
| boolean | equals (java.lang.Object other) |
| int | hashCode () |
| java.lang.String | toString () |
Static Public Member Functions | |
| static ICPMethod | of (int value) |
Public Attributes | |
| final int | value |
Static Public Attributes | |
| static final ICPMethod | Combined = new ICPMethod(0) |
| static final ICPMethod | PointToPoint = new ICPMethod(1) |
| static final ICPMethod | PointToPlane = new ICPMethod(2) |
The method how to update transformation from point pairs
The constants are static fields of this class. value holds the underlying value, so this can also hold values that aren't named constants.
| meshlib.ICPMethod.ICPMethod | ( | int | value | ) |
| boolean meshlib.ICPMethod.equals | ( | java.lang.Object | other | ) |
| int meshlib.ICPMethod.hashCode | ( | ) |
|
static |
Returns the named constant with this value, or a new instance if there's none.
| java.lang.String meshlib.ICPMethod.toString | ( | ) |
< PointToPoint for the first 2 iterations, and PointToPlane for the remaining iterations
< select transformation that minimizes mean squared distance between a point and a plane via the other point in each pair, < converge much faster than PointToPoint in case of many good (with not all points/normals in one plane) pairs < This mode only works correctly if the object that points are being projected onto has normals information (is not a point clouds without normals). < By default both objects are projected onto each other, so at least one of them must have normals, unless you change the settings < to only project one way (using MR::ICP::setFltSamples() or MR::ICP::setRefSamples()), in which case the object being projected onto < must have normals.
< select transformation that minimizes mean squared distance between two points in each pair, < it is the safest approach but can converge slowly
| final int meshlib.ICPMethod.value |