1504 
photo control, the distances between centres 
of guides 1-3, 1-15, 3-13, and 13-15 were 
taped. 
Photography was taken with a Zeiss UMK 
10/1318 camera mounted on a platform can- 
tilevered out from the base of a personnel 
basket. The basket with camera operator was 
then hoisted by crane to a height of 12m and 
positioned over the panel. A stereopair was 
obtained by rotating the jib of the crane in 
order to give a base distance of approximately 
4m between the exposures. This method of 
photography proved very satisfactory and rel- 
ative tilt of the photographs was easily ac- 
commodated during analysis. 
Initially the photographs were measured 
on a Wild A7 plotting instrument equipped 
with an EK5a coordinate recorder. Later, a 
Zeiss Steko 1818 stereocomparator fitted 
with digitisers was used to measure the same 
photographs. After taking control measure- 
ments, the following points were observed in 
the stereomodel: (1) the centre of each con- 
ductor guide as defined by the intersection of 
the string lines, and (2) at least eight points 
around the outside perimeter of each conduc- 
tor guide. 
From these measurements, the distances 
between centres of the conductor guides 
were computed. Comparisons were then 
TABLE 1. 
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING, 1975 
made between each photogrammetric meas- 
urement and that produced by conventional 
means: also, the two photogrammetric meas- 
urements were compared. These results are 
given in Table 1. Whilst the two photogram- 
metric measurements show a good consisten- 
cy (RMS error of + 1.6mm), two large discrep- 
ancies of about 10mm and 15mm occurred 
between the conventional tape survey and 
the photogrammetric survey. Both of these 
distances were subsequently measured by 
tape and the discrepancies were reduced to 
3mm and 5mm respectively. 
By using the photogrammetric measure- 
ments, the coordinates of the centres of each 
conductor guide as defined by the string lines 
and as determined from a best-fit circle pro- 
gram were compared. The distances between 
these centres are listed in Table 2 and givea 
RMS error of + 2.2mm. This indicates that it 
is unnecessary when using the photogram- 
metric method to define the centres by string 
lines. Table 2 also lists the radius of the best- 
fit circle to the eight measured points, to- 
gether with the RMS error for each tube, 
which is a useful indicator of roundness. 
SHIPS 
Over the past eight years, the Department 
of Surveying, University of Newcastle upon 
COMPARISON OF THE DISTANCES BETWEEN TUBE CENTRES (DEFINED BY 
INTERSECTING STRINGS) AS DERIVED BY PHOTOGRAMMETRY AND TAPE SURVEY FOR 
CONDUCTOR PANEL -136E, 
  
Tubes Distance by 
Distance by 
Differences (mm) 
Tape Survey (m) Photogrammetry (m) 
Wild A7 Steko 1818 
  
1 2 3 1-2 13 2-3 
1-4 2.4448 2.4394 2.4364 +5.4 +8.4 +3.0 
4-7 2.4352 2.4342 2.4346 +1.0 +0.6 —0.4 
7-10 2.4368 2.4358 2.4379 +1.0 —1.1 —2.1 
10-13 2.4400 2.4444 2.4436 —4.4 —3.6 +0.8 
2-5 2.4479 2.4325 2.4333 +154 +146 —0.8 
5-8 2.4352 2.4342 2.4333 +1.0 +1.9 +0.9 
8-11 2.4305 2.4348 2.4351 —4.3 —4.6 —0.3 
11-14 2.4321 2.4329 2.4326 —0.8 —0.5 +0.3 
3-6 2.4368 2.4361 2.4350 +0.7 +1.8 +1.1 
6-9 2.4321 2.4298 2.4328 +2.3 —0.7 —3.0 
9-12 2.4448 2.4477 2.4482 —2.9 —3.4 —0.5 
12-15 2.4368 2.4417 2.4396 —4.9 —2.8 +2.1 
1—2 2.4352 2.4316 2.4328 +3.6 +2.4 —1.2 
2-3 2.4305 2.4294 2.4305 +1.1 0.0 -1.1 
4-5 2.4384 2.4391 2.4371 —0.7 +1.3 +2.0 
5-6 2.4162 2.4140 2.4164 +2.2 —0.2 —2.4 
7-8 2.4384 2.4379 2.4396 +0.5 —1.2 -1.7 
8-9 2.4321 2.4230 2.4217 +9.1 +104 +1.3 
10-11 2.4416 2.4371 2.4377 +4.5 +3.9 —0.6 
11-12 2.4273 2.4221 2.4212 +5.2 +6.1 +0.9 
13-14 2.4305 2.4333 2.4327 —2.8 —2.2 +0.6 
14-15 24273 2.4325 2.4301 -5.2 -2.8 +2.4 
RMS error +49 +4.8 +1.6