How to implement point cloud affine registration?

### Checklist

- [X] I have searched for [similar issues](https://github.com/isl-org/Open3D/issues).
- [X] For Python issues, I have tested with the [latest development wheel](http://www.open3d.org/docs/latest/getting_started.html#development-version-pip).
- [X] I have checked the [release documentation](http://www.open3d.org/docs/release/) and the [latest documentation](http://www.open3d.org/docs/latest/) (for `master` branch).

### My Question

I am learning the point cloud registration by the [document](http://www.open3d.org/docs/latest/tutorial/pipelines/global_registration.html).

The provided demo search the transform with `rotation angle` and `translation`. But, I want to search a scale parameters.

For example, the original registration point is:


If I modify the transform matrix to:

```
trans_init = np.asarray([[0.0, 0.0, 2.0, 0.0],
[2.0, 0.0, 0.0, 0.0],
[0.0, 2.0, 0.0, 0.0],
[0.0, 0.0, 0.0, 1.0]])
```

and the visualization result is:


Then, the registration result is:


The registration result is obsolutely wrong. The reason is that the demo only conside a rigid registration rather than a affine registration.

How can I implement a affine registration with `open3d`? Any suggestion is appreciated~~~