# Projecting surface landmarks on sample model to target model in ALPACA

**URL:** <https://discourse.slicer.org/t/projecting-surface-landmarks-on-sample-model-to-target-model-in-alpaca/48170>\
**Category:** Support\
**Tags:** alpaca\
**Created:** [September 15, 2026, 2:06pm UTC](https://discourse.slicer.org/t/projecting-surface-landmarks-on-sample-model-to-target-model-in-alpaca/48170 "2026-09-15T14:06:06Z")\
**Posts on this page:** 5\
**Page:** 1

<div class="post-metadata">

**Author:** ![JYTan](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.slicer.org/jytan/32/82814_2.png) [@JYTan](https://discourse.slicer.org/u/JYTan)\
**Post date:** [September 15, 2026, 2:06pm UTC](https://discourse.slicer.org/t/projecting-surface-landmarks-on-sample-model-to-target-model-in-alpaca/48170/1 "2026-09-15T14:06:06Z")

</div>

Hi, I am new user of 3D slicer. I have generated a template mesh and surface landmarks of orbital volume segmentation and wish to project same numbers of surface landmarks from the template model to a set of target models using ALPACA in 3D slicer version 5.10.0. May I ask does same number of landmarks will be eventually projected at both the source and target models after running ALPACA? As the number of points are different between source and target pointclouds after running subsampling.. and May I ask is it possible to check the number of landmarks finally registered on both the model after running ALPACA? The tutorial I searched from community are from the previous version of 3D slicer and some functions have been removed or combined in current version. Highly appreciate if any explanation or suggestion could be provided. Thank you.

---

<div class="post-metadata">

**Author:** ![muratmaga](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.slicer.org/muratmaga/32/3622_2.png) [@muratmaga](https://discourse.slicer.org/u/muratmaga)\
**Post date:** [September 15, 2026, 3:35pm UTC](https://discourse.slicer.org/t/projecting-surface-landmarks-on-sample-model-to-target-model-in-alpaca/48170/2 "2026-09-15T15:35:05Z")

</div>

> [@JYTan](#):
>
> ish to project same numbers of surface landmarks from the template model to a set of target models using ALPACA

Thats exactly what ALPACA does.

> [@JYTan](#):
>
> May I ask is it possible to check the number of landmarks finally registered on both the model after running ALPACA?

No, need to check. Number of transferred landmarks will identical to the ones you have for your source (template) model.

> [@JYTan](#):
>
> e tutorial I searched from community

This is the official tutorial. Screenshot are small changes to the UI, but the functionality is identical to the documented workflow. [Tutorials/ALPACA/README.md at main · SlicerMorph/Tutorials · GitHub](https://github.com/SlicerMorph/Tutorials/blob/main/ALPACA/README.md)

---

<div class="post-metadata">

**Author:** ![JYTan](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.slicer.org/jytan/32/82814_2.png) [@JYTan](https://discourse.slicer.org/u/JYTan)\
**Post date:** [September 15, 2026, 4:13pm UTC](https://discourse.slicer.org/t/projecting-surface-landmarks-on-sample-model-to-target-model-in-alpaca/48170/3 "2026-09-15T16:13:04Z")

</div>

> [@muratmaga](#):
>
> No, need to check. Number of transferred landmarks will identical to the ones you have for your source (template).

Thank you for the clarification. May I ask does it mean that the number of transferred landmarks is same with the number of source landmarks used as template at the beginning, and the “run subsampling” step is functioned to generate point clouds of source and target for use to allign and capture the point of correspondence to project the landmark? What will be affected by the density of point clouds is the detail of morphology of the specimens and eventually the correspondence of the projected landmarks, instead of the number of landmarks. Is my understanding correct?

---

<div class="post-metadata">

**Author:** ![muratmaga](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.slicer.org/muratmaga/32/3622_2.png) [@muratmaga](https://discourse.slicer.org/u/muratmaga)\
**Post date:** [September 15, 2026, 4:35pm UTC](https://discourse.slicer.org/t/projecting-surface-landmarks-on-sample-model-to-target-model-in-alpaca/48170/4 "2026-09-15T16:35:11Z")

</div>

> [@JYTan](#):
>
> and the “run subsampling” step is functioned to generate point clouds of source and target for use to allign and capture the point of correspondence to project the landmark?

You probably need to to read the orginal paper to understand the details of how ALPACA works: [https://besjournals.onlinelibrary.wiley.com/doi/10.1111/2041-210X.13689](https://besjournals.onlinelibrary.wiley.com/doi/10.1111/2041-210X.13689)

Subsampling steps involves extracting sufficiently dense point cloud to do the deformable registration which is then used to transfer the landmarks on the source model.

---

<div class="post-metadata">

**Author:** ![JYTan](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.slicer.org/jytan/32/82814_2.png) [@JYTan](https://discourse.slicer.org/u/JYTan)\
**Post date:** [September 15, 2026, 4:37pm UTC](https://discourse.slicer.org/t/projecting-surface-landmarks-on-sample-model-to-target-model-in-alpaca/48170/5 "2026-09-15T16:37:49Z")

</div>

Understood. Thank you for the clarification.