# Transform FittingΒΆ

Demonstrate how to estimate rigid body motion of a point cloud

ExampleTransformFitting.java

``` 1   public static void main(String[] args) {
2       Random rand = new Random(234);
3
4       // Create a transform which will be applied to the point cloud
5       Se3_F64 actual = new Se3_F64();
6       ConvertRotation3D_F64.eulerToMatrix(EulerType.XYZ,0.5,-2,0.15,actual.R);
7       actual.T.setTo(2,3,-2);
8
9       // Create a random point cloud and transform it
10       List<Point3D_F64> fromPts = new ArrayList<Point3D_F64>();
11       List<Point3D_F64> toPts = new ArrayList<Point3D_F64>();
12
13       for( int i = 0; i < 50; i++ ) {
14           double x = rand.nextGaussian();
15           double y = rand.nextGaussian();
16           double z = rand.nextGaussian();
17
18           Point3D_F64 p = new Point3D_F64(x,y,z);
19           Point3D_F64 tranP = new Point3D_F64();
20
21           SePointOps_F64.transform(actual,p,tranP);
22
25       }
26
27       // Estimate the transform from the point pairs
28       MotionTransformPoint<Se3_F64, Point3D_F64> estimator = new MotionSe3PointSVD_F64();
29
30       if(!estimator.process(fromPts,toPts))
31           throw new RuntimeException("Estimation of Se3 failed");
32
33       Se3_F64 found = estimator.getTransformSrcToDst();
34
35       // Print out the results and see how it compares to ground truth
36       actual.print();
37       System.out.println();
38       found.print();
39   }
```