However, photogrammetry has a number of limitations
and is not always effective. For example dense conifer
forest, deciduous forest in leaf, swampland of tall
grasses and featureless terrain such as sand dunes
can be mapped in more details by ground survey
techniques. Furthermore there is no substitute for on-
site inspection of terrain features, especially in large-
scale mapping of built-up areas, where a considerable
volume of underground detail and paved surface is
encountered.
In addition, with the evolution of GIS there is a need
for descriptive information that can be acquired only in
the field, e.g., usage of building - commercial, industry,
residential or needs for data verification in the field to
ensure greater thematic and positional accuracy.
For these reason there is a great demand for field
completion in the process of producing high quality
large scale databases.
The basic objectives of field completion are:
- to select information from areas which were obscure
at the photogrammetric workstation.
- to complete thematic classification and to add
supplementary data. Some users may require the
inclusion of a specific type of information in their
geographic database (e.g., identification of electricity
network elements and collection of their attributes,
classification of buildings according to their usage and
number of floors, etc.).
- to verify that the map conforms to accuracy
requirements.
In this paper we are proposing a new approach that
will make field completion work more efficient, cost
effective and less error pron. The use of spatial
analysis techniques allows the identification of areas
to be completed, optimization of field completion
missions --routing, selection of methods and
instruments, costing, etc.-- and eventually provides the
necessary quality indicators. This has been achieved
utilizing GIS/CAD software tools.
2. INTEGRATION OF GIS, PHOTOGRAMMETRY
AND FIELD COMPLETION
The creation of a geospatial database is typically a
multistep process where GIS with its spatial analysis
functions and different types of data are involved.
Figure 1 illustrates the process and the various stages
for its creation.
We will describe how we integrate those phases to
one work flow with special attention to the new
modules.
144
Our approach concerns mainly with the last stages C,
D and E considering that the initial spatial database
has been created in the stages A and B.
A. Planning - The process starts when the client
approves the answer to his request for proposal
(RFP). At this step the requirements should fully be
analyzed and categorized. Some demands can be
achieved only with field completion, whereas others
can be satisfied with data collected by
photogrammetric means.
The planning stage also includes:
- detailed schedule with missions list and completion
time.
- designing of a photogrammetric flight configuration
- translation of the specification into detailed
instructions
- preparation of a data collection menu --table of
symbols, linetypes etc.
Customer Specification
A | |
Î i
| Requirements analysis & general working plan i
EEE T Eins ith | = I rm EL. UE
| ontrol Survey
|
|
| Photogrammet Triangulation
orientation
Database structuring and analysis
Tu.
ysiation le Len
To Mon: Work plan & Uncertainty map:
Manually input
Trees & vegetation
Relief displacement
i a ces qe Shadow areas
Lee CE NN HELLE” Different sources |
D
Special requirments.--- "_ ere:
e poc dig Field completion ; gr
Data insertio nee 5 ..... New concept
ess sea MS Standard procedure
E Ma x
” Quality estimation {1 Maybe improved
[es Next step
Figure 1 : Spatial database creation
B. Data collection - This is the "indoor" -office - data
collection stage, it is initiated and directed by the first
phase of planning and preparation. The collection of
photogrammetric data and data from other sources is
done at this stage.
Collection of data from other sources consists of two
steps:
- gathering of maps and information from the archives
of related organizations --Municipality, Electricity
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B4. Vienna 1996