Figure 2: Ned Kelly’s Armour
Image © State Library of Victoria
Unfortunately there was no leg protection, and this eventually
brought Ned undone. In the resulting shoot out, where ‘friendly
fire’ by the Police killed innocent women and children, Ned
Kelly was wounded and captured, and taken to Beechworth. He
was then moved to Melbourne, tried and hanged. This however
was not the end of the matter, the legend of Ned Kelly is as
strong or even stronger today as it was in the 19" Century.
The Data Acquisition Process.
Previous experience with the ModelMaker H40 Laser Scanner
and Faro Coordinate Measurement Arm with the Statue of
Zeus/Poseidon in Greece meant that this equipment was again
used to scan this cultural artefact. The Faro articulated arm
allows the scanning head to access difficult locations on the
artefact, offering considerably more versatility than other fixed
mount scanning systems (Da Kosta, 2000). The arm used in
this project is the Gold Version with a 10 feet reach, with a
positioning accuracy of +0.084mm/+0.0033in. The Model-
Maker scanner has a scan dimension of 40mm wide at a dis-
tance of 90mm, and an estimated precision of 0.1mm.
The ModelMaker scanner uses a projected laser stripe to gener-
ate a profile across the surface of the object under study. This
process relies on the reflection of the laser from the material,
and naturally some surfaces are far easier to measure than oth-
ers. The iron used in the armour, which over the years has been
subjected to a variety of conservation treatments, unfortunately
exhibited very poor reflectance. Even with the laser on high
power it was very difficult to obtain measurements on much of
the armour. Conventionally a spray would be used to create a
matte treatment over the surface, but in this instance this was
unacceptable. The data acquisition was slowed considerably as
a result. Many passes were made over some sections in order to
obtain complete coverage, resulting in dense point clouds
around small regions of very sparse data. The data was filtered
and repaired in the data processing stage.
In addition to the problems with the reflectivity of the surface,
on occasion various movable sections of the armour obscured
other features. In the case of the helmet, the front section is in
effect a visor which has been bolted to the cylindrical section.
The visor created an obstacle to the scanning, it obscured
important detail on the front of the helmet and it was very unsta-
portant detail on the front of the helmet and it was very unsta-
ble. Whilst it would have been preferable to have removed the
visor, this again was unacceptable to the conservation staff.
Likewise with the lappet or apron, the one attached to the front
section of the armour is not original, however it has been at-
tached for over 100 years so it too needed to remain in place
With the metal hasp obscuring detail.
Figure 3: Scanning the Armour
Figure 4: The laser on the helmet
Data Processing and Visualisation
The laser scanner provides a very dense point cloud of 3d data
points representing the surface of the object. The ModelMaker
does not record the colour of the surface, so it was necessary to
use the medium format film based images of the armour taken
as part of this documentation project by a professional photog-
rapher.
As mentioned previously the difficulty in scanning the dark iron
of the armour meant that in some places there were very dense
data points, and in others the mesh was somewhat more sparse.
The first stage in the data processing is to filter the data to re-
duce redundant points, making the data sets more manageable in
size. The second stage is to infill sections where there are insuf-
ficient points, using the options in the ModelMaker software.
For the armour, this was used along the edge of the panels
where it was very difficult to get a laser reflection.
The next stage in the processing was to join the inside and out-
side scans of the armour panels. During the scanning process
wherever possible sections of both the inside and outside sur-
—546—
