Last modified: January 07, 2021

This article is written in: 🇺🇸

Data Types and Structures

VTK uses 3D geometries, including points, lines, polygons, and volumes. It handles images and volumetric data for 2D and 3D visualization. It works with scalar, vector, and tensor fields for complex data representation. Supports structured and unstructured grid types for various spatial data layouts. Includes support for time-series data and hierarchical datasets.

vtkDataObject

I. Base class for all data objects in VTK

II. Stores data and metadata including

• Geometric Information: Encompasses shape, position, and size of the data elements. Essential for defining the spatial representation of the data.
• Topological Information: Concerns the connectivity of data elements. It reveals the relationships between elements, like how points are linked to form lines or polygons in a mesh.

III. Common subclasses include

1. vtkImageData: Handles regular, rectilinear grid data.
2. vtkRectilinearGrid: Manages grid data with varying spacing.
3. vtkStructuredGrid: Deals with structured data points in 3D space.
4. vtkPolyData: Specializes in representing polygonal data.
5. vtkUnstructuredGrid: Ideal for representing complex, irregular grid data.

|
|--- vtkDataObject
     |
     |--- vtkDataSet
          |
          |--- vtkPointSet
          |    |
          |    |--- vtkPolyData
          |    |
          |    |--- vtkStructuredPoints (vtkImageData)
          |    |
          |    |--- vtkStructuredGrid
          |    |
          |    |--- vtkUnstructuredGrid
          |
          |--- vtkRectilinearGrid
          |
          |--- vtkCompositeDataSet
               |
               |--- vtkHierarchicalBoxDataSet
               |
               |--- vtkMultiBlockDataSet
               |
               |--- vtkHierarchicalDataSet
               |
               |--- vtkOverlappingAMR
               |
               |--- vtkNonOverlappingAMR

vtkImageData

• Description: Represents a regular grid with fixed topology and uniform spacing between points.
• Applications: Ideal for volumetric data like images or 3D scalar fields.
• Examples: CT and MRI scans, where vtkImageData stores pixel values at each grid point for 3D visualization.

vtkRectilinearGrid

• Description: A regular grid with fixed topology but non-uniform spacing between points.
• Applications: Suitable for data with varying resolution, like in climate or terrain elevation data.
• Examples: Climate models with varying altitude resolution, or terrain data with resolution changing with slope.

vtkPolyData

• Description: Represents a dataset comprising points, vertices, lines, polygons, and triangle strips.
• Applications: Ideal for surface meshes and 3D models.
• Examples: Models of 3D objects (e.g., vehicles, buildings), terrain surfaces (mountains, valleys).

vtkStructuredGrid

• Description: Curvilinear grid maintaining fixed topology.
• Applications: Best for data on curvilinear coordinate systems.
• Examples: Fluid flow simulations around objects, finite volume method simulations with a fixed cell division.

vtkUnstructuredGrid

• Description: An irregular grid with flexible topology, capable of containing various geometric primitives.
• Applications: Suitable for complex geometries and adaptive meshes.
• Examples: Finite element method simulations with complex, dynamic domains, models of intricate 3D geometries (human brain, aircraft wings).

Structured vs Unstructured Grids

Multiblock Dataset

• Description: A collection of multiple datasets, organized hierarchically.
• Applications: Handles complex, hierarchical data structures effectively.
• Examples: Multi-domain simulations (each domain as a separate dataset), multi-resolution data (datasets at different detail levels).

A vtkMultiBlockDataSet organizes datasets (or blocks) hierarchically. Each block can be a vtkDataSet subclass or another composite dataset, allowing versatile dataset organization. For instance, a vtkMultiBlockDataSet might contain blocks like vtkPolyData, vtkStructuredGrid, and even another vtkMultiBlockDataSet. This structure is invaluable in large-scale simulations where data is segmented into blocks representing different simulation areas or system components, enabling VTK to manage complex, multi-part datasets coherently.

Choosing the Right Data Structure

• Consider the geometry, topology, and resolution of your data
• The choice of data structure affects memory usage, processing time, and ease of use
• Some data structures can be converted to others, but not always without loss of information