angle allowable. It is good, however, to be confronted with the 
worst conditions possible because this helps us to appreciate all 
the problems associated with updating topographic data using 
the SCO. 
5.3 Software Selection 
In 1983, New Brunswick chose the CARIS software to produce 
its digital topographic infostructure. Since then, CARIS 
software developments (Derenyi 1991)(Derenyi and Teng 1992) 
added the soft orthophoto functionality. Subsequently Universal 
Systems Ltd. incorporated and enhanced the software 
development as a standard package in the CARIS products 
(Masry, Mayer and Clarke 1994). Although the first objective 
of the project was to evaluate the soft orthophoto technology as 
a tool for digital data revision, it was also an opportunity to 
gain some experience with the orthorectification module of the 
CARIS software. 
5.4 Hardware Selection 
In a large project, one may select a powerful computer system 
to optimize the human and hardware resources. In an 
evaluation project, the hardware resources are not critical. 
Both Northeast Exploration Services and Geomacadie Services 
used a similar SUN SPARC Station IPX, equipped with a 2.6 
GB hard disk, a 2.3 GB tape drive, a 16 MB RAM, and a 16- 
inch standard colour monitor. 
EXPERIMENTAL SITE 
  
5.5 Spatial Extent 
As described earlier, the philosophy is to use existing 
photography to minimize the revision cost of the digital 
topographic base. The existing photography was flown at the 
experimental forest site. The Department of Natural Resources 
and Energy’s experimental forest site is located in Central New 
Brunswick (see Figure 5-1). The extent of the site is equivalent 
to a map at a scale of 1:50 000. The New Brunswick 
topographic system is a breakdown of the 1:50 000 NTS. The 
1:50 000 encompasses twenty-five maps at a scale of 1:10 000. 
Therefore, the aerial photography covered twenty-five 1:10 000 
maps. The file for the Village of Stanley was the one chosen 
because the area provides the most changes in topographic 
feature and height variations (160 metres), a very good area for 
312 
checking the orthophotographic and mosaicking processes. The 
aerial photography lines fall near the northern and southern 
edges of the map (see Figure 5-2), which enables us to check 
the mosaicking in the direction of the flight lines as well as 
between them. Also, the lateral mosaicking gives us a clearer 
idea of the software’s efficiency in handling radiometric 
corrections between the flight lines. 
FLIGHT LINES 
  
  
  
  
  
  
1 Y HR ) T ce - D. re th med 
\ : ez : 7 j ANS | " c7 St cx 225 pi a PS 
A M. sum die ATS 2e T T XC — 
Figure 5-2 
5.6 Phases 
5.6.1 Input 
Three intrinsic data bases (Figure 5-3) are essential for 
execution of the project: digital aerial photography (DAP), 
digital topographic data (DTD) and digital elevation models 
(DEM). 
A large number of scanners in the industry are capable of 
scanning aerial photography. The cost of such scanners 
varies considerably. It is very likely that the geometric and 
optical qualities of these scanners are generally in 
proportion to their cost. Quite possibly, there are very 
good scanners at a reasonable price, and it is also quite 
possible to perform geometric calibrations on a scanner. In 
a production environment, the choice of an appropriate 
scanner is essential. In our situation, the project's variables 
had to be minimized. For this reason, it was decided to do 
the scanning on a precision scanner (Zeiss PhotoScan PS 
1). 
The film (negative) was sent to the company responsible for 
the scanning (Triathalon mapping Services) so as to be sure 
that the diapositives could be printed at the optimal density 
for scanning, that they would be free of ink or grease spots, 
fingerprints, scratches, and that the emulsion would be on 
the proper side of the photographic medium. 
The scanning was done at a resolution of 15 um, which 
corresponds to a resolution of 0.45 metre on the ground. 
The use of 24 bits in the scanning made it possible to obtain 
16.7 million colours. The data was provided on an 
Exabyte-type magnetic medium in the Tag Image File 
Format (TIFF). 
The first reason for these specifications was to capture all 
  
the de 
secon 
two p 
paper 
applic 
the re 
contr: 
eleva: 
nine ] 
to the 
5.6.2 
The | 
were 
IPX : 
in it: 
prepr 
The | 
size). 
The : 
Red, 
possi 
data » 
Hue . 
chan; 
mani; 
5.6.3 
The f 
data : 
edito: 
fiduc 
The