Voxel Remesh

The Voxel Remesh tool reconstructs a surface mesh by converting it to a volumetric representation and then extracting a new isosurface. This voxel-based approach fundamentally differs from traditional remeshing by treating the surface as a volume boundary, enabling powerful mesh repair and regularization capabilities.

Overview

Voxel remeshing operates through a three-stage process:

1. Voxelization: The surface is converted to a 3D voxel grid where each voxel is classified as inside or outside the object
2. Distance field computation: A signed distance field is calculated from the voxel boundary
3. Isosurface extraction: A new mesh is generated by extracting the zero-level isosurface

This volumetric approach offers unique advantages over surface-based remeshing:

• Automatic topology repair: Self-intersections, holes, and non-manifold geometry are automatically resolved
• Noise removal: Small surface defects are naturally smoothed during voxelization
• Shell unification: Overlapping or intersecting shells are merged into a coherent surface
• Consistent normals: The extracted surface has correctly oriented normals throughout

Accessing the Tool

Navigate to the Surface ribbon tab and locate Voxel Remesh in the Edit section. Select one or more surfaces before activating the tool.

Voxelization Methods

Two methods are available, each optimizing for different priorities:

Accurate Method

The accurate method prioritizes geometric fidelity by using a finer voxel grid and more precise distance field calculations. This approach:

• Preserves fine geometric details better
• Uses more memory and processing time
• Produces results closer to the original surface

Use the accurate method when geometric precision is paramount, such as for manufacturing, measurement, or detailed anatomical models.

Fast Method

The fast method optimizes for processing speed using efficient algorithms that may sacrifice some geometric detail:

• Significantly faster processing
• Lower memory requirements
• Suitable for preview or interactive workflows

The fast method is ideal for initial exploration, large datasets, or situations where processing time is a constraint.

Voxel Spacing Configuration

The voxel spacing determines the resolution of the volumetric grid and directly impacts both the quality of the result and the computational requirements:

X, Y, Z Spacing

You can specify independent voxel dimensions along each axis:

• X spacing: Voxel dimension in the X direction (mm)
• Y spacing: Voxel dimension in the Y direction (mm)
• Z spacing: Voxel dimension in the Z direction (mm)

Smaller spacing values produce finer resolution and better geometric preservation but require more memory and processing time. The relationship is cubic—halving the voxel spacing increases memory requirements by a factor of eight.

Uniform Spacing

Enable Uniform spacing to use the same voxel size in all directions. This is recommended for most applications as it:

• Produces isotropic mesh quality
• Simplifies parameter selection
• Avoids directional artifacts in the result

When uniform spacing is enabled, adjusting any axis value automatically updates all three dimensions.

Choosing Appropriate Spacing

The optimal voxel spacing depends on the smallest features you need to preserve:

Feature Size	Recommended Spacing	Application
Large features only	1.0 - 2.0 mm	Quick repair, simplification
Standard detail	0.3 - 0.5 mm	General purpose remeshing
Fine detail	0.1 - 0.2 mm	Detailed anatomical models
Very fine detail	0.05 - 0.1 mm	High-precision applications

As a guideline, the voxel spacing should be approximately 1/3 to 1/2 the size of the smallest features you want to preserve.

Smoothing Option

Enable the Smooth option to apply post-processing smoothing to the extracted surface. This additional step:

• Reduces staircase artifacts from voxelization
• Produces smoother surface appearance
• Slightly reduces geometric accuracy

Smoothing is particularly beneficial when using larger voxel spacings, as it mitigates the blocky artifacts that can result from coarse voxelization.

Use Cases

Mesh Repair

Voxel remeshing is exceptionally effective for repairing problematic meshes:

Self-intersecting surfaces: When triangles penetrate each other, voxelization naturally resolves the intersections by treating the surface as a boundary rather than individual triangles.

Non-manifold geometry: Edges shared by more than two triangles, or vertices where the surface topology is ambiguous, are automatically corrected during isosurface extraction.

Holes and gaps: Small holes in the surface are often filled during voxelization if they are smaller than the voxel spacing.

Boolean Operation Preparation

Surfaces intended for Boolean operations must be watertight and topologically valid. Voxel remeshing can prepare problematic meshes for Boolean operations by ensuring:

• Closed, watertight surfaces
• Consistent normal orientation
• Removal of self-intersections

Merging Multiple Shells

When a surface contains multiple overlapping or adjacent shells that should be unified, voxel remeshing naturally merges them into a single coherent surface based on the volumetric inside/outside classification.

Technical Considerations

Memory Requirements

Voxel remeshing requires memory proportional to the volume of the bounding box divided by the cube of the voxel spacing:

$$\text{Memory} \propto \frac{V_{bbox}}{s^3}$$

where $V_{bbox}$ is the bounding box volume and $s$ is the voxel spacing.

For large objects or fine voxel spacings, memory requirements can become substantial. Consider:

• Using the fast method for large objects
• Starting with coarser spacing and refining as needed
• Processing different regions separately if memory is limited

Detail Loss

Voxelization inherently limits the minimum feature size that can be represented. Features smaller than approximately 2-3 voxels may be lost or distorted. This behavior is often beneficial for noise removal but may be problematic for surfaces with intentionally fine features.

Watertight Requirement

The voxelization process requires surfaces to be reasonably close to watertight for proper inside/outside classification. Surfaces with large holes or open boundaries may produce unexpected results, as the algorithm cannot reliably determine the interior volume.

Comparison with Standard Remesh

Aspect	Remesh	Voxel Remesh
Approach	Surface-based triangulation	Volume-based reconstruction
Topology repair	Limited	Comprehensive
Feature preservation	Edge preservation option	Limited by voxel resolution
Speed	Generally faster	Depends on voxel count
Memory	Proportional to triangle count	Proportional to volume/spacing³
Best for	Quality improvement	Repair and unification

Choose standard remeshing when the input mesh is topologically valid and you primarily want improved triangle quality. Choose voxel remeshing when you need to repair topology, merge shells, or remove self-intersections.