rbal
map"
data
sily
°way-
gian
lesy-
A COMPARISON OF THE EFFECTIVENESS OF FOUR DATA CAPTURE PRODUCTION LINES
Paper #1260, Poster Session 3, Thursday August 6th. 1992
S.P. Goel
Survey of India (SOI)
: J.E. Drummond
International Institute for Aerospace Survey and Earth Sciences (ITC), The Netherlands
A.M. Tuladhar
International Institute for Aerospace Survey and Earth Sciences (ITC), The Netherlands
PURPOSE:
The Survey of India (SOI) is considering building a National Topographic Database model of the Indian
landscape to the standards and level of detail associated with India’s 1:250,000 map series. The project
reported here proposes that data will be
captured regionally, and to this end
it is necessary to find
the most cost effective data input means for implementation at regional centres. This paper summarises a
joint SOI/ITC investigation carried out on four
cost and high cost scanning and vectorizing,
KEYWORDS: Map data capture. Comparison of methods.
Screen’ Digitizing.
1. INTRODUCTION
This paper reports on an investigation to identify
an appropriate spatial data capture method for
contributing to building a National Digital
Database for India. This spatial component can be
referred to as a National Topographic Database. As
in any similar endeavour, it is necessary to
identify the users of the database and their needs
before designing the data-model, identifying the
sources, and selecting the sources and methods of
data capture. However once the sources have been
identified, it is possible to design objective
tests to choose the most appropriate method(s) of
data capture. In India it has been planned that
the spatial component of the National Digital
Database will be subdivided into databases
representing real world entities at the levels of
detail found on maps at the scales 1:4,000,000;
1:1,000,000; 1:250,000; 1:50,000; and 1:25,000
(Agrawal, 1989). The following table summarises the
1991 situation regarding some of these Indian
series, when considering conventional maps.
Series % Complete Number of sheets
complete (approx)
1:25,000 30% 7000
1:50,000 100% 5000
1:250,000 100% 300
Table 1 - Status of Indian map series
In this investigation it was proposed to start
building the spatial component of a National
Digital Database using the 1:250,000 series. The
reasons for this were: 1) it was a potentially
manageable task from which experience could be
gained in capturing and handling large data sets;
2) although of low resolution it would enable the
rapid future production of updated 1:250,000 scale
maps, which are currently only produced by the
manual generalisation of 1:50,000 maps; 3) with
only approximately 300 sheets a national database
could be built quite quickly; and 4) three users
potential data production lines (manual digitizing, low
: and 'heads up'
these in terms of the effort required to achieve a certain quality of end product.
or 'on-screen' digitizing) and compares
Time trials. Manual Digitizing. Scan Digitizing. 'On-
197
(Forest Survey, Soil Survey, and The Survey of
India (SOI)) have expressed an interest in this
scale (Stefanovic, 1990a).
In the 1980's digital Cartography was introduced
to the SOI. By 1991 in the SOI's Modern
Cartographic Centre there was a SysScan Integrated
Digital Map Production System, in its Automated
Cartography Cell there was another Computer
Assisted Mapping installation, in its Digital
Mapping Centres were Integraph Map Production
Systems, and in its Survey Training Institute
several Integraph MicroStation-PCs were used. A
core of personnel were already experienced in
these systems, and so it was decided that any
future production lines (in terms of database
building) should not deviate far from these
existing systems. However during this
investigation it was realised that database
building and maintenance could be decentralised to
the regional offices of the Survey of India; such
decentralisation leads to a requirement for low
investment cost systems. ;
The situation outlined in the preceding paragraph
influenced the four production lines tested in
this work. The production lines involved: 1)
manual digitizing and interactive editing in the
Intergraph PC-MicroStation environment; 2)
scanning, vectorizing, and interactive editing in
the SysScan environment; 3) low-cost scanning
(AGFA), vectorizing (SysScan environment), and
interactive editing (SysScan environment); and 4)
scanning, ‘on-screen’ digitizing and interactive
editing in the Digisys environment. (Digisys is a
low-cost MS-DOS based integrated raster/vector
cartographic editing system developed by the Dutch
company Reprocart bv. It takes input from Scitex
and other scanners and produces colour separated
plotting files for Scitex laser film
printer/plotters (Korver, 1991).) These four
production lines will be further discussed in
section 3 of this paper.
