Documentation of 'boofcv.alg.geo.robust.DistanceSe3SymmetricSq' Java class
DistanceSe3SymmetricSq
boofcv.alg.geo.robust

Class DistanceSe3SymmetricSq

  • All Implemented Interfaces:
    DistanceModelStereoPixels<Se3_F64,AssociatedPair>, DistanceFromModel<Se3_F64,AssociatedPair>


    public class DistanceSe3SymmetricSq
    extends java.lang.Object
    implements DistanceModelStereoPixels<Se3_F64,AssociatedPair>

    Computes the error for a given camera motion from two calibrated views. First a point is triangulated from the two views and the motion. Then the difference between the observed and projected point is found at each view. Error is normalized pixel difference squared.

    error = Δx12 + Δy12 + Δx22 + Δy22

    Error units can be in either pixels2 or unit less (normalized pixel coordinates). To compute the error in pixels pass in the correct intrinsic calibration parameters in the constructor. Otherwise pass in fx=1.fy=1,skew=0 for normalized.

    NOTE: If a point does not pass the positive depth constraint then a very large error is returned.

    NOTE: The provided transform must be from the key frame into the current frame.

    • Method Summary

      All Methods Instance Methods Concrete Methods 
      Modifier and Type Method and Description
      double computeDistance(AssociatedPair obs)
      Computes the error given the motion model
      void computeDistance(java.util.List<AssociatedPair> associatedPairs, double[] distance)
      Computes the distance a set of points is from the model and saves the results in the provided array.
      void setIntrinsic(double cam1_fx, double cam1_fy, double cam1_skew, double cam2_fx, double cam2_fy, double cam2_skew)
      Specifies intrinsic parameters See comment above about how to specify error units using intrinsic parameters.
      void setModel(Se3_F64 keyToCurr)
      Sets the model parameters.
      • Methods inherited from class java.lang.Object

        equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
    • Constructor Detail

      • DistanceSe3SymmetricSq

        public DistanceSe3SymmetricSq(TriangulateTwoViewsCalibrated triangulate,
                                      double key_fx,
                                      double key_fy,
                                      double key_skew,
                                      double curr_fx,
                                      double curr_fy,
                                      double curr_skew)
        Configure distance calculation.
        Parameters:
        triangulate - Triangulates the intersection of two observations
      • DistanceSe3SymmetricSq

        public DistanceSe3SymmetricSq(TriangulateTwoViewsCalibrated triangulate)
        Configure distance calculation.
        Parameters:
        triangulate - Triangulates the intersection of two observations
    • Method Detail

      • setIntrinsic

        public void setIntrinsic(double cam1_fx,
                                 double cam1_fy,
                                 double cam1_skew,
                                 double cam2_fx,
                                 double cam2_fy,
                                 double cam2_skew)
        Specifies intrinsic parameters See comment above about how to specify error units using intrinsic parameters.
        Specified by:
        setIntrinsic in interface DistanceModelStereoPixels<Se3_F64,AssociatedPair>
        Parameters:
        cam1_fx - intrinsic parameter: focal length x for camera 1
        cam1_fy - intrinsic parameter: focal length y for camera 1
        cam1_skew - intrinsic parameter: skew for camera 1 (usually zero)
        cam2_fx - intrinsic parameter: focal length x for camera 2
        cam2_fy - intrinsic parameter: focal length y for camera 2
        cam2_skew - intrinsic parameter: skew for camera 2 (usually zero)
      • computeDistance

        public void computeDistance(java.util.List<AssociatedPair> associatedPairs,
                                    double[] distance)
        Description copied from interface: DistanceFromModel
        Computes the distance a set of points is from the model and saves the results in the provided array.
        Specified by:
        computeDistance in interface DistanceFromModel<Se3_F64,AssociatedPair>
        Parameters:
        associatedPairs - Set of points which are to be evaluated.
        distance - Where model distance is stored.

DataMelt 3.0 © DataMelt by jWork.ORG

You see the box below because you did not login.