# How to find the upper-direction of mesh data ( idea , task, reference anything )

**URL:** <https://discourse.slicer.org/t/how-to-find-the-upper-direction-of-mesh-data-idea-task-reference-anything/27737>\
**Category:** Development\
**Tags:** python\
**Created:** [February 10, 2023, 6:51am UTC](https://discourse.slicer.org/t/how-to-find-the-upper-direction-of-mesh-data-idea-task-reference-anything/27737 "2023-02-10T06:51:03Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![dsa934](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.slicer.org/dsa934/32/17907_2.png) [@dsa934](https://discourse.slicer.org/u/dsa934)\
**Post date:** [February 10, 2023, 6:51am UTC](https://discourse.slicer.org/t/how-to-find-the-upper-direction-of-mesh-data-idea-task-reference-anything/27737/1 "2023-02-10T06:51:03Z")

</div>

Operating system: window 10  
Slicer version: 5.2.1

I have several mesh data files (.ply), and the orientation of objects in these files is always random.

like this :

 ![image](https://us1.discourse-cdn.com/flex002/uploads/slicer/original/3X/7/5/757ac353c96ed48389f043def4cfef74819c12f0.png)

With the goal of having all meshes have the same ‘upper-direction’, I tried the following ideas.

1. calculate mesh’s center point

2. get ‘inner-mesh’ from kD-tree

3. Find the point normals of the ‘inner-mesh’ and sum them all.

```auto

# make functions that find inner mesh 

# set mesh's center point

mesh_center_pos =[0,0,0]
node_in_scene = slicer.mrmlScene.GetNodesByClass('vtkMRMLModelNode')

for value in node_in_scene:
 node_name = value.GetName()

crown_node = getNode(node_name)
mesh = crown_node.GetMesh()

# make kdtree for mesh data
points = mesh.GetPoints()
k_Dtree = vtk.vtkKdTree()
k_Dtree.SetNumberOfRegionsOrMore(4)
k_Dtree.BuildLocatorFromPoints(points)

# cal mesh's center point 
for value in range(mesh.GetNumberOfPoints()):
 mesh_center_pos[0] += mesh.GetPoint(value)[0] 
 mesh_center_pos[1] += mesh.GetPoint(value)[1] 
 mesh_center_pos[2] += mesh.GetPoint(value)[2] 

mesh_center_pos[0] /= mesh.GetNumberOfPoints()
mesh_center_pos[1] /= mesh.GetNumberOfPoints()
mesh_center_pos[2] /= mesh.GetNumberOfPoints()

# current pos = [28.750410911221405, -4.0782514881915715, 26.817471450103564]
slicer.modules.markups.logic().AddControlPoint(mesh_center_pos[0], mesh_center_pos[1] , mesh_center_pos[2])

inner_mesh_id = vtk.vtkIdList()

k_Dtree.FindClosestNPoints( 3000, mesh_center_pos, inner_mesh_id)

inner_mesh_pos = [] 

# get nearest position's id 
for value in range(inner_mesh_id.GetNumberOfIds()):

 inner_mesh_pos.append( inner_mesh_id.GetId(value))

# draw nearest position via makrupsFiducialNode
for value in inner_mesh_pos:
 closet_pos = mesh.GetPoint(value)
 slicer.modules.markups.logic().AddControlPoint(closet_pos[0], closet_pos[1] , closet_pos[2])

```

However, the kD\_tree.FindClosestNPoints() function has a problem because it requires directly specifying the number of points close to the reference point.

Therefore, I want to unify the case where the direction of the mesh data object is random. Is there a way to find the uppder-direction of the ojbect mesh?

---

<div class="post-metadata">

**Author:** ![lassoan](https://sea2.discourse-cdn.com/flex002/user_avatar/discourse.slicer.org/lassoan/32/13_2.png) [@lassoan](https://discourse.slicer.org/u/lassoan)\
**Post date:** [February 15, 2023, 3:57am UTC](https://discourse.slicer.org/t/how-to-find-the-upper-direction-of-mesh-data-idea-task-reference-anything/27737/2 "2023-02-15T03:57:30Z")

</div>

If the shape is always similar to your sketches then the “up” vector is the vector that connects the _center of the oriented bounding box_ to the _center of gravity_. You can get both these center positions from Segment Statistics module.