Digitization on a budget: A Smithsonian open-source automated pipeline
Whether it is using AR to visualize Ikea furniture in your home or hunting down a Snorlax in Pokemon Go!, 3D technologies are becoming more commonplace in everyday life. Matching this trend, tools used to create content for these 3D applications are also becoming more common. In the museum and archive space we see this reflected through increased 3D digitization of collections. While many methods are available, photogrammetry is the preferred technique for many institutions, likely due to its low barrier to entry, quality of color capture, and data durability. Though photogrammetry capture devices have become cheaper and more accessible, the processing of this data often relies on expensive proprietary solutions, or a disparate combination of open-source, and not always user-friendly, tools that can be a blocker for individuals and institutions with a limited budget.
This how-to session aims to demystify the process taking you from image set to final 3D model deliverable. We will cover how to implement and use a full photogrammetry data processing pipeline that consists of only open-source tools (no software budget required!) and automates many tasks to reduce the expertise needed. This session will not discuss photogrammetry capture itself but will focus on taking existing photogrammetry data and turning it into 3D models usable for a range of purposes.
We will cover the following key steps of the photogrammetry processing pipeline:
- Image pre-processing
- Photogrammetry processing
- Camera alignment
- Model Reconstruction
- Model post-processing
- UV Unwrap
- Texture projection
- Web packaged deliverable creation
* This how-to session will not cover capturing a photogrammetry dataset
11089 [pods name="Paper" template="user_block" Where="_mw_paper_proposal_id="][/pods]