Wikipedia:Guide for creating 3D models for Wikipedia

This is a guide to producing and using 3D models of real objects and environments for Wikimedia projects.

3D models are now supported on Wikimedia Commons, derivatives from 3D models including those using photogrammetry are also useful captured as images or videos:

• to show objects from directions it is not possible to take a photograph.

• to create videos exploring the object, something that may be difficult to do, e.g. fly throughs of buildings or of small objects.

Advantages

Timelapse video of a hyperboloid object print (made of PLA) using a RepRap "Prusa Mendel" 3D printer for molten polymer deposition

3D models on Wikipedia may allow people to understand environments and objects more thoroughly and in new ways including:

• Objects that exist on different scales to us or those that are not visible, very small objects like proteins and very large environments and structures like volcanoes.

• Allow virtual access to sites that they have no other way of visiting either through physical distance or restricted access.

• Could be used in conjunction with GLAM initiatives. A way museums and other institutions of all sizes could create better documentation of their collection for basically free and offer a more rich way of viewing objects on and not on display.

• An amazing addition for articles about living creatures.

• Whilst some models do not capture the whole of objects and may leave holes they potentially offer far more information than photographs.

• Could be used as files to print the objects using a 3D printer either by mail order or a home printer e.g a Reprap. This has massive education potential in organisations like museums being able to print replicas of objects, in schools making educational models and at home, especially for mechanical models and people with vision problems who understand the world through touch.

Creating accurate 3d models

Photogrammetry

Photogrammetry is the process of taking several photographs of an object or environment and using them to generate a 3d model of the object. Using free programs that are simple to use, well documented and don’t need a fast computer or special equipment to run accurate 3d models can be generated. A higher resolution (10MP+) is helpful although it is dependent on contrast of images and number of images taken. There are two sorts of 3d model that can be created using free software:

Wire-frame models are a series of lines that connect to make a shape that can have a surface texture, using photogrammetry to produce these models can result in holes and bumps where information is missing but give a detailed representation of objects. 123d Catch is available as a desktop application, smartphone app (iOS) and a web application. Photographs are uploaded to a server where the model is calculated which can then be downloaded.

• Gibraltar 1 Neandertal skull 3d wireframe model, created with 123d Catch.
• 3d model of Horatio Nelson bust in Monmouth Museum, Wales

Point clouds are points that map the surface of an object, they are often better for complicated objects and environments (especially those with trees and other irregular shaped objects) as gaps and lack of information are often less noticeable.

• 
  Monmouth Castle point cloud, created with Photosynth.
• 
  The Monmouth Castle point cloud seen from above, no photos were taken from this angle.

Correcting errors

It is possible to correct errors and remove extraneous points on models with a free 3D modelling program e.g Blender.

Other sources of data to create 3D models

Libraries of open license datasets e.g LIDAR measurements could be used to generate 3d models e.g NASA and other US federal sources.

Preexisting libraries of models

There are already many sites offering free open licence models that could be added to Wikipedia including:

• 123D Catch, e.g this model of a Tyrannosaurus Rex.

• Photosynth: including this model of Arles Amphitheatre.

• Thingiverse, where people have already started to document museum collections in 3d under a CC-BY-SA license.

• Libre3D, solely focused on free/open license models: www.Libre3D.com.

• Models could be created using photographs that already exist, they would work best using photos with similar lighting conditions.

Technical progress on adding 3d models to Wikipedia

Some progress is already being done including the wonderful work by User:emw who has created an extension to allow interactive 3D models of proteins and DNA on Wikipedia, large libraries of 3d models exist for molecules that could imported.

Open projects do exist for displaying 3d models online by Mozilla and Google.

Inclusion in Wikimedia projects is discussed at bugzilla:1790.