This project studied whether existing three-dimensional (3D) digital photo imaging technologies can be easily adopted by historic sites and collections for documenting and monitoring the condition of buildings and objects. It also aimed to determine whether workflows could be established that would result in consistent and scientifically reliable 3D condition data capture.
The Problem: For over a century, conservators have relied on 2-dimensional (2D) photography to document and monitor the condition of heritage sites and objects. When a sudden and shocking visual indicator appeared (e.g., a sudden crack in a surface), conservators took notice of what had actually been thousands of slow, tiny, incremental condition changes (e.g., tiny fissures and failures in slowly aging materials). Conservators then photographed and monitored the area to try to determine how rapidly the deterioration was advancing. Two factors limited our ability to detect and monitor deterioration: our ability detect slow and small “micro” changes, and our ability to compare what humans see using 3D stereo vision and what we recorded in 2D photographs.
Digital 3D images or surrogates allow us to see and measure 3D volumes and contours and they allow us to use a computer to detect and highlight very small and slowly occurring changes. Detailed 3D models can be made using laser scanning and 2D digital photography (Photogrammetry, Reflectance Transformation Imaging –RTI- and Structured light). Each technology has condition-imaging strengths and weaknesses. In each, the finer the resolution – that is, the smaller the area captured by either a laser point reflection or a digital camera sensor pixel, the larger the data set for an entire object will be. In practice, the larger the object you want to image – an entire historic house or landscape versus a single door or wall, for example – the lower your resolution or detail will need to be or the more expensive and customized your assembly computer will have to be. Laser scanning equipment uses proprietary laser projectors, collection cameras, software and digital file formats. The files are huge – tens to hundreds of gigabytes each, and require very specialized computing hardware and software packages.
The Technological Solution: This study found that two 3D digital photographic technologies –Highlight RTI and Photogrammetry – promise to change the condition-detecting and monitoring paradigm in important and beneficial ways:
- 3D photo images allow us to record and study surfaces and volumes in ways that mimic the stereo view of a real-time human examiner.
- Because specular, color and texture data is digital, a viewer can selectively highlight and study details by removing the color data, changing the virtual light source and reflectivity or colorize contour levels.
- Photography can be done by anyone with a basic knowledge of digital camera operation and a lap-top computer. Laser scanning engineers are not required.
- Photography uses an ordinary, consumer-professional grade, digital camera. Expensive and complex laser scanning equipment is not required.
- Capture photographs are the same, open-format 2D photos you have always taken –RAW, TIFF and JPEG. The capture data is not a proprietary laser-scan file format that requires special, brand-name software to use and view.
- The user owns their own images and controls their naming, archiving, meta-data and digital management protocols.
- Assembled 3D, computational meshes, solids, and RTI images and their viewers are open-source software – there are no hidden steps and every transformation of the data is documented and visible.
- Because the capture photos are regular, open-format digital images, 3D images can be assembled with whatever future enhanced versions of 3D assembly software happen to come along. You are not dependent upon the life of one assembly software or manufacturer.
- Capture photography can use the best digital camera the heritage site can afford. While 21 megapixel, full frame CMOS sensors with prime lenses and RAW capture capability give the highest resolution, less expensive APS-C sensor cameras that have manual capture modes (allowing the user to fix the lens aperture and focus) give reliable 3D condition documentation.
We created an interactive, web-platform blog site that documented our evolving working methodologies, equipment and resulting images. We specifically targeted the historic preservation field and publicized the project blog, videos and workflows using historic preservation groups on Linked-In, Facebook and Twitter social media outlets. While we anticipated somewhere near 3000 views of the website content during the course of the project, we were astonished by how rapidly an international audience, heritage preservation audience for the project was developed:
9000 views were from the USA and Caribbean.
4000 views were from Europe
700 views were from Central and South America.
500 views were from Asia (China does not report by Chinese government regulation)
400 views from Australia and New Zealand
325 views were from Canada
200 views were from the Middle East
70 views were from India
35 views were from Africa
15,230 views, total