4. CONCLUSIONS 
The determining factors in deciding what interval 
to use between the REGULAR points are: 
1) Purpose of the contour map. 
2) Relief of the terrain in the area. 
3) Map scale and contour interval. 
If high levels of accuracy are required in specific 
areas of the project, additional geographic points 
can be added to increase the accuracy of the data 
extracted from the DTM in that area. But system 
errors and human errors may cause irregualar errors 
in the finished contours. Therefore a general 
guidance formula is needed to guide the operators 
toward increasing work efficiency and the reliabil- 
ity of the DTM data. 
Concerning the reliability of the DTM created 
contours, the suggestion is to establish a proce- 
dure for digitization of geographic elements for 
the DTM files. It is important to digitize the 
verification contours on a different layer of the 
graphic file after all of the geometric elements 
are digitized. These verification contours can then 
be compared with the contours generated with DTM 
data. 
From the comparative analysis between the contours 
generated from the DTM (TTN) files and the digi- 
tized contours by manual stereocompilation, we 
derived the following conclusions: 
1) Contours created by manual stereocompilation 
are usually higher than the contours generated 
by DTM files. Refer to Fig.6, the most of dif- 
ferences between 25 FT spacing DTM and manual 
stereocompilation are minus. 
2) Reduction in the density of REGULAR points 
will slightly reduce the accuracy of the 
contours if the amount of REGULAR points is 
still adequate to produce the desired contour 
interval in the map. (see Fig.6) 
3) An increased amount of REGULAR points will 
create contours with smoother curves, i.e. 
contours expressing more detailed relief. 
4) BREAK lines play an important role to repre- 
sent abrupt irregularities in the terrain. If 
BREAK lines are omitted, it may lead a signif- 
icant errors in the resulting model. 
(see Fig.2 A) 
If the REGULAR points are reduced, only rela- 
tively minor errors will result. 
5) Digitization of the geographic BREAK lines can 
compensate for a reduction in the density of 
REGULAR points. This is a feasible method to 
enhance the DTM accuracy. 
6) Unexpected human errors caused the Standard 
Error from the 40-FT DTM file to be greater 
than the Standard Error generated from the 50- 
FT DTM file. (see Table.l & Fig.6) 
The stereoplotter has been reset for the 
digitization of 40-FT DTM after the digitiza- 
tion of 50-FT DTM. 
7) The accuracy of the contours generated with 
DTM data using REGULAR points at the 50 FT 
interval were acceptable for highway design. 
8) More experiments are necessary to set up 
detailed specifications for the application of 
DTM data. 
9) Comparing contours generated from different 
point intervals and from manual compilation is 
one of the primary methods to verify the 
overall accuracy of the DTM data. 
DTM applications are not accepted by some profes- 
sional engineers, because they are reluctant to 
adopt the results generated from DTM technology. 
Lack of literature supporting the technical viabil- 
ity of computer software is a major issue hindering 
widespread acceptance of DTMs within the profes- 
sion. 
The accuracy of the DTM data affects the quality of 
the results generated from DTM applications. The 
above method of comparing contours from different 
criteria can persuade engineers that the results 
generated from DTM data are as good as the results 
expected from traditional photogrammetric approach- 
es. Therefore, eventually more advanced DTM appli- 
cations will be supported and utilized by profes- 
sional engineers. 
5. ACKNOWLEDGEMENT 
The author wishes to express his appreciation to 
Mr. William Blank for his suggestions and support, 
to Mr. Stanley Kijek for his digitization, to Mr. 
Dave Keilly for his revisions, and to Buchart-Horn 
Inc. for their support. 
6. REFERENCES 
+ References from BOOKS: 
a) 1989. Engineering Site Package (ESP) and 
Transportation Design Package (TDP), User's 
Guide. Intergraph Corp. pp 7-3:7-9. 
b) 1991. InSite/InRoads Reference Guide. 
Intergraph Corp. pp GL-3:GL-9. 
846