Proceedings Template

1.1 Challenges facing Muse= ums

· Increasing amount of content (physical items), and at the same time= a decreasing storage space for the items

· Memory institutions are relatively few in number, and unequally distributed which causes inequities in what is chosen to be stored, and wha= t is accessible.

· Access to content is severely limited due to travel costs. And again there are inequities due= to many groups having little to no funding.

· Many antiquities are too fragile to travel, or to allow repeated handling and exposure.

· Handling of culturally sensitive materials may not be permissible or appropriate.

· Content items or sites are lost to natural hazards (floods, fires, volcanoes, earthquakes), theft, warfare, or economic development.

= Sites may not be available

=

1.2 Advantages to Virtual Museums

There are many advantages to digitizing content, both individual items and exhibi= ts (even the entire museum). Hav= ing a digital copy of items or exhibits allows them to be accessed by anyone, at = any time, from any place. It allows any number of people access at the same time. It preserves a nearly complete rec= ord of the object, which can be accessed without damaging the original. 3D digitization and display potent= ially solve all the challenges listed above. The advantages that virtual museums provide are listed below:

· Imagery and spatial measurements are mechanically recorded and not subject to human interpretation.

· Everything can be recorded, in complete detail, in their original setting, with limited human effort.

· Precise, repeatable measurements are made that are equally or more precise than human measured ones.

· Objects can be viewed virtually and virtually dissected for study f= or any amount of time, with no cost or damage to the content.

· Morphological comparison of related material is facilitated through qualitative visual comparisons or quantitative shape-based comparisons.

· The system is non-invasive, in that it does not touch or affect the samples or site. This is impo= rtant for conservation.

· Because the data is digital it can be conveniently archived and made available anywhere, anytime, to anyone 24/7/365.

· Different interfaces, or visualizations, can be provided depending = on the observer, their task, and the material. For instance a scholar may desire a shape comparison display while a museum visitor may wish to simply browse t= he different exhibits.

= ;

1.3 Equipment and Methods<= o:p>

The equipment used for scanning the scenes is the DeltaSphere-3000 (3rdTech, Chapel Hill, North Carolina<= /st1:State>, http://3rdtech.com/DeltaSphere.htm). It combines laser rangefinder technology, professional digital photography, and state-of-the-art computer graphics software in a portable instrument that can be mounted on a photography tripod (Figure 1). The DeltaSphere-3000 employs an embedded time-of-flight laser rangefinder--a de= vice that measures the distance, or range, to any point the laser hits. It uses a rotating mirror to scan a vertical slice, and a rotating motor in the base that rotates around the vertical axis for the next slice (see Figure 1). This process is repeated = until the specified field of view has been covered. The coordinate system used is much like the latitude/longitude system on the surface of the earth, or the azimuth/elevation coordinates used in surveying.

Computer control of the internal positioning motors at the base, and the revolving mirror allow the= DeltaSphere to automatically scan a complete room or scene with the laser rangefinder.<= span style=3D'mso-spacerun:yes'> The default setting of 13.33 samples/degree is appropriate for scanning rooms or large scenes for virtual museum digitizations. Using t= his setting

= Figure 1 DeltaSphere 3D Scene Digitizer

the scanner records the range and position of several million sample points for distances up to 40ft (12.2m) from the scanner in less than 20 minutes. The acquired set of sample points = can be automatically converted to a simple 3D model. This model can be rotated, scaled, and displayed from arbitrary viewpoints. It can be used as input to other software packages for creating realistic 3D images and animations. Finally, range data from multiple = scans can be combined to create a single 3D model. For example, scans of multiple rooms, or multiple parts of t= he same room can be combined to create a complete detailed model.

The second step is to = use a professional digital camera to capture the color image data for the scene.<= span style=3D'mso-spacerun:yes'> We used a Fuji FinePixS2Pro with non-fisheye lens AF Nikkor ED= 14mm f/2.8D for the examples in this paper.&nbs= p; The captured color digital images are correlated with the laser range finder spatial points. This a= llows the generation of very realistic views of the 3D scene from any angle. An example of a static 2D rendered= image from one viewpoint is seen in Figure 2. While most range finder based digitization systems can provide 3D scene views, the images are generally of lower quality. The 3D environ= ment recorded by the DeltaSphere 3000, however, is of very high quality, nearly indistinguishable from a photograph of the scene as seen in figure 2. While static 2D images can be used= to view the digitized 3D environment, it is more effective to view it using real-time viewing applications that display the 3D scene on a 2D computer display maintaining some of the 3D visual cues (lighting, shading, obscurat= ion, stereo (if stereo viewing glasses are used), user controlled changing of viewpoint). This supports a v= irtual reality experience where users can actually feel as if they are in the muse= um, as opposed to seeing photographs of the museum. Techniques that photographically capture scenes, or that do not integrate the color image textures with the range finder data cannot provide such visualizations. Figure 3 shows the actual sample d= ata points underlying the visualization seen in figure 2. The output of the pro= cess of combining the color texture from the digital photographs with the laser range finder sample points is a VRML format data file. This is a standard format for text= ure mapped polygons, and supported by most 3D viewing applications.<= /span>