— Ws 
35 
-——va 
29 
rative 
2s be 
“ferent 
> been 
sively 
wd for 
3 logue, 
1e. The 
ocedure 
able 1. 
or each 
A few 
ed for 
od ca 
should 
of the 
t. ihe 
he cen- 
diam- 
e gen- 
special 
ial. to 
tead of 
sons of 
uracies 
within 
alibra- 
re not 
of the 
As a 
-tradi- 
es per- 
ombined 
on 3x 
rammet- 
oratory 
oulos & 
produc- 
us the 
  
  
  
  
  
  
  
  
  
  
  
  
  
METHOD CAMERA INSTRUMENT PLOTTING EDITING 
ANALOGUE Wild P31/100 Technocart D Graphical by hand 
ANALYTICAL Hasselblad/80 ADAM MPS-2 Automatic AutoCAD 
LOW COST—-HYBRID Wild P31/100 |Digitizer+S/W Automatic AutoCAD 
Table 1 
DTM of the vessel's hull was produced and 
it was later interpolated along prede- 
fined lines in order to produce the final 
plots. 
One of the main problems with all methods 
is the uncertainty with which the various 
lines of interest are being defined on 
the hull, in the photogrammetric stereo— 
model. With the analogue method, the in- 
strument may be driven along specific 
planes, provided of course, that the co- 
ordinate system of the control is ident- 
ical with that of the final shipbuilding 
plan. Even then a relatively experienced 
operator may be faced with difficulties. 
With the analytical method, things are 
theoretically more favourable. However, 
it proved equally difficult to drive the 
instrument along predefined planes, thus 
determining their section with the 
object. Special off-line interpolation 
stage was necessary for both methods, in 
order to achieve the required results, as 
described above. In addition the lack of 
detail points on the surface of the hull 
presents another problem to the observer. 
Since it is impossible to either paint or 
project onto the hull some kind of tex- 
ture, the operator is required to proceed 
very carefully. 
    
In all cases the lines produced needed 
some kind of smoothing. This may either 
be done graphically with special drawing 
aids, or computationally. A special CAD 
Programme was developed for this purpose. 
As input it requires co-ordinates of 
observed points on the sections and it 
then determines the appropriate curve to 
fit (usually of third order). This is 
determined by the type of the vessel, its 
dimensions and the position of the sec- 
tion. 
An interesting problem is to attempt the 
survey of such a vessel while it is in 
the water, when a portion of the ship's 
hull is underwater. Two are the obvious 
solutions. In the case of clear waters 
the photography could be taken from out- 
side and special two-media algorithms 
should be employed. Alternatively  under- 
water photography with specialised  ins- 
trumentation should be employed. Both 
alternatives will be tested in the course 
of a major research project, which will, 
hopefully, start in the near future. 
  
n—-— 7777 
o 1 2m 
Figure 3 
On the other hand, the final method, 
Proved easier and faster in any respect. 
As the DTM could be produced in any co- 
ordinate system, no special care should 
be taken for the determination of the 
control. Hence simple distance measure- 
ments may be used for the analytical ori- 
entation of the stereopair, thus making 
the fieldwork easier, simpler and more 
accessible by à non-photogrammetrist. 
Careful specification of the plane para- 
meters produces more reliable sections of 
the object. The only drawback of this 
method is the - temporary - lack of ste- 
reoscopy for the observations. 
4. CONCLUDING WORDS 
Since the main issue in this discussion 
is the photogrammetric methodology, it is 
proposed that simple instrumentation com- 
bined with simple measurements and simple 
software may lead to the required result. 
Of course it has been clear from the ex- 
periments that the best way to handle the 
problem is to produce a DTM of the ves- 
sel's hull and produce the required  sec- 
tions from it.