masking tape is feasible where the surface of pipe is smooth
and low temperatures exist. It is relatively simple to apply
masking tape. However, high temperature or surface roughness
(e.g., rusted surfaces) might be a barrier to employing this
technique.
4. 3DMODELLING PROCESS
Three sets of scanning data were joined using a cloud-based
registration method, which was introduced recently (Fall,
2002). Unlike in control/tie point-based registration, the
registration can be done with points without setting out control
targets around model space. This application can not only
eliminate the time and cost for setting artificial targets but also
reduce the difficulties of setting up targets on or around hardly
accessible area. Each scan set is registered with minimum of
three common points. These points can be located anywhere,
but an edge of an object or very distinctive points should be
chosen to track them easily in the other scan sets. The resulting
scanning cloud mosaic captures a part of the table and piping
objects.
To model each pipe point-clouds of each object was separated
by selecting window and "cleaning points" was necessary to
eliminate unnecessary points, which can be mainly noise and
points of adjacent object such as a table in this experiment.
After point-clouds were selected the fitting cylinder was
constructed. The diameter of cylinders was determined based
on fitting object. Diameter of each object is compared with
‘true’ diameter measured by a caliper in 0.01mm. In the case
of galvanized iron, ceramic, and aluminum the measured values
from scanning data were within 2 to 4 mm differences, which
were compared with ‘true’ diameter. However, the difference
was 17mm with 38mm PVC pipe, and this was unacceptable.
5. Conclusions
According to the observation of coverage angles, PVC and
stainless steel have the smallest coverage angles while brass
and ceramic pipes have much wider angles. In order to form a
model properly at least 20 to 25% of coverage angle is
recommended.
The effects of dulling spray application were quite
distinguishable and the results showed 50% to 60% increases in
coverage angles of point clouds compared to the “as-is”
condition. Masking tape was tested on the aluminum surface
and the area of scanning point clouds was increased. As a
result, the accuracy of the measurement of an object was
improved. The procedure of applying these media was
relatively simple and efficient in an industrial environment, and
their application is reliable for object measurements and
modeling processes.
The diameter measurements of piping objects based on 3D
model fitting showed the measurements were close to the ‘true
value’ in the case of ceramic, aluminum, and galvanized iron.
However, measurements on some materials such as PVC and
Stainless steel were unacceptable range.
6. References
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B5. Istanbul 2004
Axelsson, P., 1999. Processing of laser scanner
data—algorithms and applications. ISPRS Journal of
Photogrammetry & Remote Sensing, 54, pp. 138-
147.
Balzani, M., Pellegrinelli, A., Perfetti, N., Russo, P.,
Uccelli, F., Tralli, S., 2002. CYRAX™ 2500 Laser
Scanner and G.P.S. Operational Flexivility: From
Detailed Close Range Surveying, to Urban Scale
Surveying. ISPRS Commission V, Symposium 2002,
Corfu, Greece.
Feng, H-Y, Liu, Y., Xi, F., 2001. Analysis of
digitizing errors of a laser scanning system.
Precision Engineering, Journal of the International
Societies for Precision Engineering and
Nanotechnology, 25, pp. 185-191.
Johansson, M., 2002. Explorations into the
Behaviour of Three Different High-Resolution
Ground-Based Laser Scanners in the Built
Environment. ISPRS Commission V, Symposium
2002, Corfu, Greece.
Lichti, D.D., Gordon, S.J., Stewart, M.P., 2002.
Ground-Based Laser Scanners: Operation, Systems
and Applications. Geomatica, Vol. 56, No. 1, pp. 21-
33.
a
di
e
um Jo £ ml a
5 7. = 2M
" Qo $ "aft.
o PE
Eom m E.
ng EN dt
g 'n_ 00 & 543 26
on Ppl ur >
CP a Pa 5 Ca
hy uo P8 on
lie vy / Be 3
ys sit 2 ^
PEE ns Ar
o " um i
VS a 05 ga
a£ 0 AT a
0e ^
aa ff 5
ve a pe
cn
x vs s ofa T
o &
LU 8
