Parametrization#

The Parametrization tab provides tools for fitting and working with models.

Sphere
#sphere
Ellipse
#ellipsoid
Cylinder
#cylinder
RBF
#rbf
Mesh
#mesh
Curve
#curve
Sample
#sample
To Cluster
#to-cluster
Remove
#remove
Merge
#merge-meshes
Volume
#volume
Repair
#repair
Remesh
#remesh
Project
#project

Parametric Fitting#

Fit basic geometric shapes to point clouds:

Sphere#

Fits a sphere using least squares optimization:

  1. Select a cluster with spherical shape

  2. Click Sphere to fit the model

  3. The fitted sphere appears in the Models section

Ellipsoid#

Fits an ellipsoid using eigenvalue decomposition and least squares optimization:

  1. Select a cluster with ellipsoidal shape

  2. Click Ellipsoid to fit the model

  3. The fitted ellipsoid appears in the Models section

Cylinder#

Fits a cylinder using PCA and iterative refinement:

  1. Select a cluster with cylindrical or tubular shape

  2. Click Cylinder to fit the model

  3. The fitted cylinder appears in the Models section

Non-Parametric Fitting#

RBF (Radial Basis Function)#

Creates smooth, non-parametric surface models through radial basis function interpolation. Ideal for complex, non-parametric shapes that can be represented as height fields, i.e. an open membrane section.

  1. Select a cluster with surface-like structure

  2. Click RBF

  3. Configure interpolation direction:

    • xy: Surface as function of x,y coordinates

    • xz: Surface as function of x,z coordinates

    • yz: Surface as function of y,z coordinates

  4. Click OK to create the interpolated surface

Mesh#

  1. Select a cluster with sufficient point density

  2. Click Mesh

  3. Choose reconstruction method:

    • Alpha Shape: Convex hull with alpha parameter control

    • Ball Pivoting: Robust surface reconstruction for structured data

    • Cluster Ball Pivoting: Ball pivoting with automatic parameter determination

    • Poisson: Watertight surface reconstruction

  4. Configure method-specific parameters:

    Alpha Shape Parameters:

    • Alpha: Controls shape complexity (higher = coarser features)

    • Scaling Factor: Mesh resampling resolution

    • Distance: Threshold for inferred vs. measured vertices

    Ball Pivoting Parameters:

    • Radii: Ball radii for reconstruction (comma-separated, e.g., “5,3.5,1.0”)

    • Downsample: Thin input point cloud to core points

    • Smoothing Steps: Pre-smoothing iterations

    Poisson Parameters:

    • Depth: Octree depth (higher = more detail)

    • Samples: Minimum points per octree node

    • Pointweight: Interpolation weight of input points

  5. Set repair parameters:

    • Elastic Weight: Controls mesh elasticity (0 = strong anchoring)

    • Curvature Weight: Controls curvature propagation

    • Volume Weight: Controls internal mesh pressure

    • Hole Size: Maximum hole area for automatic filling

  6. Click OK to generate the mesh

Note: Mesh quality depends on point cloud density and noise levels. For noisy data, increase smoothing steps. For sparse data, reduce the number of neighbors.

Curve#

Fits spline curves of requested order to sequential control point data. Good for creating smooth curves from hand-drawn paths:

  1. Create control points using drawing mode:

    • Press Shift+A to enter curve drawing mode

    • Click to place control points in sequence

    • Press Enter to complete the curve

    • OR select an existing cluster with linear structure

  2. Click Curve

  3. Configure spline parameters: - Order: Spline degree (1=linear, 3=cubic, 5=quintic)

  4. Click OK to fit the curve

Sampling Operations#

Sample#

Creates point clouds from fitted models:

  1. Select one or more models in the Object Browser

  2. Click Sample

  3. Configure sampling parameters:

    Sampling Method:

    • Points: Generate specified number of points

    • Distance: Generate points with specified average spacing

    :Parameters:**

    • Sampling: Number of points or point spacing value

    • Offset: Normal-direction offset from surface (useful for particle picking)

  4. Click OK to generate sample points

To Cluster#

Converts model vertices to point cloud format:

  1. Select one or more models in the Object Browser

  2. Click To Cluster

  3. Model vertices are automatically added as new clusters

Note

For most models, both vertices and computed normals are preserved in the conversion.

Remove#

Deletes selected models:

  1. Select one or more models in the Object Browser

  2. Click Remove or press Delete

  3. Selected models are permanently removed

Mesh Operations#

Volume#

Creates meshes from volumetric data using marching cubes:

  1. Click Volume

  2. Select a volume file (MRC, MAP, EM, HDF5)

  3. Configure meshing parameters:

    Algorithm Settings:

    • Simplification Factor: Reduce triangle count by this factor

    • Workers: Number of parallel processing threads

    • Close Dataset Edges: Create closed meshes at volume boundaries

    Processing Method:

    • Uses sharded marching cubes algorithm for large volumes

    • Splits volume into manageable chunks for parallel processing

    • Merges submeshes and applies quadratic edge collapse simplification

  4. Click OK to generate meshes

Note

Optimized for large datasets with automatic memory management and parallel processing.

Repair#

Fixes mesh topology issues and fills holes using Leipa triangulation and fairing:

  1. Select mesh models to repair

  2. Click Repair

  3. Configure repair parameters:

    Optimization Weights:

    • Elastic Weight: Mesh smoothness (0=anchor to original, 1=free movement)

    • Curvature Weight: Preserve or modify curvature

    • Volume Weight: Internal pressure (positive=inflation, negative=shrinkage)

    • Boundary Ring: Optimize n-ring vertices around boundaries

    Hole Filling:

    • Hole Size: Maximum hole area to fill (-1=fill all holes)

  4. Click OK to repair meshes

Remesh#

Improves mesh quality and adjusts triangle density:

  1. Select mesh models to remesh

  2. Click Remesh

  3. Choose remeshing method:

    Edge Length:

    • Edge Length: Target average edge length

    • Iterations: Number of optimization passes

    • Mesh Angle: Preserve edges above this angle threshold

    Vertex Clustering:

    • Radius: Clustering radius for vertex merging

    Quadratic Decimation:

    • Triangles: Target triangle count

    Subdivide:

    • Iterations: Number of subdivision passes

    • Smooth: Use smooth Loop subdivision vs. simple midpoint

  4. Configure method-specific parameters

  5. Click OK to remesh

Use Cases:

  • Edge Length: Create uniform triangle sizes for simulation

  • Vertex Clustering: Quick mesh simplification

  • Quadratic Decimation: High-quality mesh reduction

  • Subdivide: Increase resolution for detailed modeling

Project#

Projects point clouds onto mesh surfaces using ray casting:

  1. Select exactly one mesh model (target surface)

  2. Select one or more point cloud clusters (sources to project)

  3. Click Project

  4. Configure projection settings:

    Projection Method:

    • Cast Normals: Use point normal vectors for ray casting

    • Invert Normals: Reverse normal direction

  5. Click OK to perform projection

Results:

  • Creates new point clouds with projected coordinates

  • Generates updated mesh with projection points integrated

  • Preserves original data while adding projected versions

Merge#

Combines multiple meshes into a single object:

  1. Select two or more mesh models in the Object Browser

  2. Click Merge

  3. Meshes are automatically combined into a single mesh

  4. Original meshes are removed, replaced by the merged result

Next Steps#

Continue to the Intelligence tab to learn about advanced features like HMFF and membrane segmentation.