Appendix B – Application of techniques to different types of specimens

Plan coordinate and seam line position prior to digitizing and mark them. Keep the limited range of the digitizer arm in mind! When aiming for a closed loft, remember to digitize start and end points! Closed curves may lead to artifacts, usually along the seam line or in places where neighboring curves are of significantly different lengths or their distance and direction varies greatly. The same is true if closed curves overlap or shift at the seam line. Here, using the 'Simplify: Rebuild...' option of the 'loft' command dialog box can reduce or remove the problem. Alternatively, the curves can be edited prior to lofting via the 'rebuild' command, and/or by editing their 'control points'.

Small bones up to 10 cm maximum length tend to be hard to digitize for a number of reasons. They shift position easily, drawing curves on or around them at relatively regular intervals without overlap alone is hard, and ensuring that these curves accurately reproduce small surface features makes the task of drawing curves free-handed too cumbersome. Using the fossil holder reduces the first problem. Still, care must be taken not to put pressure on the specimen with the tip of the digitizer as this may shift the bone. As it is hard to place a small bone on the holder stably while still being able to reach both upper and lower surfaces, we recommend first digitizing only the upper half with concentric closed curves in a horizontal plane (Figure 15.1), then flipping the bone over and repeating the procedure for the other half. This requires a coordinate set C_set1 placed on the circumference of the bone (Figure 15.2).

The problem of drawing curves properly - especially closed curves - can be solved by extensive marking of the bone. For this, much or all of the bone is covered in painter's adhesive tape or masking tape and the seam line and curves are drawn onto it (Figure 15.2). It is also possible to only mark partial curves or markers for curve passage over radially placed strips of tape if the bone is too large for complete wrapping. It is of paramount importance to select a tape that can be removed completely without damaging the bone. The resulting curve set can be lofted normally (Figure 15.3).

For medium sized elements (up to around 90 cm maximum length), curves should be placed roughly perpendicularly to the long axis of the bone, with the seam line on the lower side. The specially constructed holder again proved best, as it keeps the specimen from shifting while allowing easy access. Usually, it will not prevent access to important features and curves placed directly to the left and right of the fittings are sufficient to give an accurate surface. If not, it is necessary to shift the bone a few centimeters and recalibrate the digitizer before adding curves into the gaps. See Figure 6 for an example of this setup.

Drawing closed curves requires that it is possible to reach around the specimen on both sides down to the seam line. For very large specimens, especially relatively slender longbones like tibiae and radii of sauropods, it is advisable to place the specimen close to vertical by leaning it against a solid structure, with a cut-to-fit styrofoam spacer to create a gap in between. Some bones, especially sauropod longbones, rest stably in this position on their articular ends. If necessary, the lower end can be secured with wooden wedges and small sandbags or in a sandbox. Thus, bending moments on the shaft are minimal, reducing the risk of damage. The digitizer may have to be set up on a wooden box or a chair in front of the bone to achieve a sufficient vertical reach; care must be given to stability. For large specimens, a forklift or similar adjustable tool can be used. This allows digitizing the top end and shaft with curves sub-parallel to the ground; then the fossil must be turned upside down, the digitizer recalibrated and the bottom end added. The seam line here runs down the 'back' of the bone, opposite the digitizer and operator during the first part, but facing the digitizer during the second part. When using early versions of Rhinoceros®, care must be given to either digitize the curves in the opposite direction after rotating the specimen (seam line now at front) or change curve direction in Rhinoceros® before or during lofting.

Bones that are too large in circumfence for closed curves or encased in supports that make reaching around them impossible can be digitized by drawing sets of open curves. Each curve in a set ends at the starting point of a curve from another set. The matching curves are then merged into one closed curve. Two or more curves together must fully surround the bone the way a closed curve would. This requires extensive marking of the bone to minimize overlap and shift between touching curves (Figure 16). It is advisable to place the contact line between curves at places with low curvature (i.e. flat bone surfaces) to minimize artifacts and leave a very small gap between curves. The partial curves then must be joined into closed curves using the 'match' and 'join' commands. Then, a closed loft can be made over the entire bone.

Usually, it is necessary to rebuild matched curves using the 'rebuild' command to avoid artifacts, as the control points will be unevenly distributed along each curve. An additional advantage of rebuilding curves is the ability to somewhat reduce the number of control points, making both the final file smaller and editing via control points simpler. This must be executed carefully, as too large a reduction in control point numbers can induce significant shape changes in the curves.

Some bone shapes cannot be represented by one loft over closed curves. In these cases, it is possible to digitize parts of the bone, loft surfaces over these parts, and then combine them into one body in Rhinoceros®. Examples are the articular heads of sauropod ribs, where one set of closed curves cannot easily be drawn around both heads. Digitizing as large a part as possible with curves in one direction, then adding a second set of curves for the missing parts is usually the best option, although it is sometimes preferable to aim to minimize the suture length between the parts.

Several points must be kept in mind when planning a composite body:

- The separate parts must together contain the entire surface or volume of the specimen when using open surfaces or closed surfaces, respectively; otherwise there will be gaps or holes. If holes cannot be avoided, they should be in flat areas of the bone, where filling them up in Rhinoceros® is easy and carries a low risk of error.
- Boolean operations on bodies (3D volumes) are much easier to conduct than joining operations on surfaces.
- The separate lofts must overlap enough or contact each other exactly to allow joining the bodies in Rhinoceros®.
- The overlap area should not contain repeat intersections between surfaces/bodies as these tend to confuse Rhinoceros® and make merging impossible. It is helpful to close lofted bodies by placing the end point inside the bone by manually adding it in Rhinoceros®. This guarantees overlap with only one section contact.
- It is advisable to digitize a closed curve along the contact line of the parts to be joined. This facilitates creation of a cutting plane, which allows cropping the parts to an exact fit. Ideally, the contact line allows the use of a flat plane for cutting.
- To create holes in a body (e.g., foramina in ischia or coracoids) it can be easier to digitize and loft an air volume and subtract it from the main body by a Boolean operation than splitting the main body into parts and joining them.
- Remember to set the tolerances tight enough, otherwise cutting and joining surfaces may not work properly.

Using open curves and multiple lofts: A closed curve should additionally be digitized along each suture line to facilitate closing each open loft into a closed body before joining the adjacent bodies.

Using closed curves and multiple lofts: A closed curve should additionally be digitized along each suture line to facilitate cropping both adjacent bodies to avoid problems when joining them.

When it is not possible to digitize one closed curve around the end of the surface it is possible to digitize it in parts, and then join them into one curve using the 'match' command.

These methods often fail for joining surfaces, due to problems with joining cropped surfaces. We strongly recommend using multiple closed lofts (bodies) and uniting them with Boolean operations.

Ideally, complex shapes should be digitized using point clouds. If that is not possible or desirable, the above procedure can also be used in a variant that delivers a polygon mesh instead of a NURBS object as the final result: Digitizing takes place with separate curve sets, then all surfaces are 'meshed', after which the meshes can be joined in Geomagic®.

Any bone too small to rest stably on the holder and/or with a shape too complex to be easily digitized by the above method for small bones using closed curves should be placed either on clay/PlayDoh® or in a sandbox. The 'Digsketch' command allows collection of point clouds. Care must be taken to ensure that the distance between closest points on the same surface is always significantly smaller than the distance to points on any other surface; otherwise artifacts will be produced during meshing. 'Meshfrompoints' produces a polygon mesh that can be edited via the 'deletemeshfaces' and 'fillhole' commands. Reducing the mesh count in Rhinoceros® is usually not advisable, as the errors incurred are significant. Both meshing and editing of the mesh are better carried out in Geomagic® (see Appendix D). This is especially true for filling gaps in the mesh.

Theoretically, Geomagic® also supports the Microscribe 3D digitizer. For unknown reasons, we were not able to make the digitizer connect to the program correctly. Despite this technical problem, Geomagic® can be a helpful tool during digitizing of point clouds. It is possible to collect data with the digitizer in Rhinoceros® while the other program is displayed on the screen (see Appendix E). Therefore, one can digitize points in Rhinoceros®, intermittently mesh them in Geomagic®, and then go on collecting data to fill the gaps shown in the mesh in Rhinoceros® while the mesh in Geomagic® is actually visible on the screen. Due to the far better ability to selectively collect data where it is needed, this process can save up to 30% of the digitizing time. Also, errors will show up earlier, allowing faster correction.

Rugosities, damaged or discolored areas, articular surfaces or muscle attachment scars can be marked on the surface by drawing a closed curve around them. This curve can later be projected onto the lofted surface. While this data will usually be lost when exporting the 3D bodies into other programs, some formats allow retention. Also, a very thin body can be constructed from the curve via the 'extrude' command; this will be preserved in all formats. Rhinoceros® allows changing the color of these markings to make them more conspicuous. Alternatively, the area can be covered with a point cloud, from which a polygon mesh can be created.

Here, an explanation is given for combining two parts; a higher number can be combined by repetition. There are two methods for combining data from two specimens. It is possible to select at least three landmarks, which are present on both objects, and digitize them as coordinates for a coordinate system, then digitize the specimens after calibrating to these landmarks. This works well if there are e.g. small foramina or sutural contacts with ontogenetically stable positions present on the bones. Otherwise, the landmarks can be digitized as points, point cloud or NURBS curve data on both objects, and the data sets can be aligned later. Optimally, only one set of closed curves is needed, with the last curve of one specimen subparallel to the first of the second specimen. If only a contralateral element is available to complete a partial specimen, it is possible to digitize one part including the landmarks and then mirror it before digitizing the other element.

If no landmarks are present, it is possible to just digitize the specimen without calibrating from one to the other, then copy and paste them into one file and move one until it appears to fit the other. This process works well if there are cristae or spinae present, or any other feature on the bone that allows easy 3D orientation of the separate parts. For both methods it is possible to scale ('scale' command) one part isometrically to fit the other. If a large size difference exists, it may be necessary to scale the element differently for various axes to compensate for the allometric growth common in vertebrates, which is possible with the 'scale2D' and 'scale1D' commands. For this, the bone should be aligned with the coordinate axes of Rhinoceros®. In any case, a note should be added to the file indicating the composite nature of the model.

Next Section