Digital morphology: The Smithsonian's biovisualization lab

C omputer-based visualization is revolutionizing many fields of science by suggesting new questions , creating new forms of data, presenting new ways to publish results, and increasing accessibility to the public through the World Wide Web. There are numerous examples of the alluring power of this three-dimensional (3-D) technology as it can objectively capture meaningful information and produce novel images of scientific specimens. The Smithsonian BioVisualization Lab was established to produce research quality digital (virtual) specimens for scientific study and education. One of our principle objectives has been to adopt a technology that could produce highly accurate digital files for taking exact measurements and for creating accurate computer-generated images of specimens. Researchers at the BioVisualization Lab have worked in collaboration with the Digital Research and Imaging Lab of the Mississippi State University to develop approaches and to establish In-ternet access that will enable others to share work and specimens on-line. The beginning of the Smithsonian BioVisualization Lab grew out of proof-of-concept experimentation with laser scanners and digital data conducted at other facilities in the late 1980s. The apparatus for magnetic resonance imaging and scanners for computer-assisted tomography and positron emission tomography are among the better-known examples of the types of tools used to create 3-D imagery (for reviews, see Hartwig and Sadler, 1 and Dean. 2) These tools have been shown to have interesting applications to anthropology, such as the reconstruction of skulls from multiple pieces. 3 The centerpiece of the BioVisualiza-tion Lab is a high-resolution laser digitizer, a point-scanning instrument. The device generates an array of x, y, z coordinates that record the geometry of an object as it spins or translates below the laser beam. The data array becomes a series of parallel contour lines that are later translated into a more refined polygonal geometry. Data collection is automated by setting various machine parameters such as point interval and laser beam focus, and by setting protocols that define a target window and calibrate the vertical distance between a point on the target surface and the laser head. When the laser fires, it triggers an internal reading of each point in an x, y plane. The reflection of the laser beam is simultaneously registered on a pair of charge-coupled device (CCD) sensors to enable triangulation of the height measurement (z coordinate value) of each datum. As the coordinate points record the geometry of the object's surface, software is used …