International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B2. Istanbul 2004
from Laser Scan Company — Lamps2. The following aspects
were considered in the implemented testing:
eLinear elements and areas for some classes should not be
interrupted;
elt is not allowed any element duplication;
e Hierarchical classes must organize coincident elements;
e Element connectivity forbidden or allowed;
e Every solid object must have closed polygons, even when
interrupted by other element; :
eAutomatic simbology for some classes with proximity
restriction.
2 2.14 Corrective Action During Vectorization Process
During vectorization, all corrective actions are related to the
operator and to the action that must be reconsidered. A pre-
interpretation can be considered as a corrective action for the
vectorization process. This can be seen, for example, in the case
of an ortophoto generation using automatic DTM and control
points generated from a map to provide a document for terrain
evaluation and object identification in the field. As a result, the
field surveyors will provide a digital document already codified
in agreement with the vectorization specification. Such a
procedure can eliminate the so called terrain interpretation
procedure and may reduce misinterpretation — during
vectorization what will lead to a gain of productivity.
22.12 Conformity Testing
To finalize, the POCS shall test some representative sampling
of the product in order to verify if they are in agreement with
the pre-established specifications. To validate the final printed
product we recommend the use of a quality procedure based on
the Quality Management from Ordinance Survey-UK.
2213 Documentation
Considering that Digital Photogrammetric Systems can store
results as ASCII files, the PQCS documentation generation is
facilitated. Two types of documents are needed in the
vectorization process:
* A report attesting the reached quality;
eA procedure documentation to clarify the data use and
how they have been validated. It is important to
consider the issue of an instruction manual or a user
manual.
At the end of each production process a proper documentation
must be issued. That documentation must be revised in order to
validate the precision and the accuracy attainted. They shall
also be presented in the final report. Some other information, as
described below, can be also included in the final report:
e Date of the applied quality test;
e Corrections and modifications applied after testing
e[ndications on how the new quality information was
verified.
3. VALIDATION OF THE PQCS
To evaluate the PQCS functionality and to validate all the
quality parameters proposed, a series of practical experiments
of quality control was applied in the following photogrammetric
procedures: scanning, interior orientation and aerotriangulation.
3.1 Scanning
In order to obtain images with high quality, a pair of
diapositives images were scanned using a DSW200 Scanner
from Leica Geosystems.
After scanning every image was submitted to a visual quality
control in a computer monitor. Also, some image regions were
analyzed particularly in zoom mode for tonality quality control.
In that case, it was verified for example the occurrence of some
dust dots in the image, as presented in Figure 3.
5t Tuc
Figure 3
The dust dots were caused by dust in the scanner glass support.
It was also verified that even after a high control cleaning and
with temperature and humidity control, a residual dust might be
found making the process vulnerable to even a strict control. In
that case, only an adequate documentation and a report of
corrective actions can help to obtain high level of quality.
If no action had been taken during scanning, the problem would
be detected only in the final process, which could be, for
instance, during ortophoto generation. In case of high quality
requirement, the work could be rejected
3.2 Interior Orientation
For interior orientation parameters quality testing, a series of
images from different companies were tested. All images were
oriented using SOCET SET Digital System from BAE Systems.
All measurements are presented in the Table 2.
798
Company Photo Camera Quantity | Pixel Size | RMS | [Pixel]
Scale of [um] [um]
images
A 1:35.000 RCIO 2 21 19,8 0,99
B 1:30.000 RC 10 2 25 23.9 0,96
B 1:30.000 | RMK TOP 4 22 9,2 0,37
cC 1:35.000 RCIO 2 21 19,8 0,99
D 1:8.000 RC8 2 20 10,3 0,52
E 1:8.000 RCIO 2 24 13,3 0,56
E 1:8.000 RCIO 2 24 18,7 0,78
F 1:8.000 RMK TOP 5 25 8,9 0,41
Table 2
The practical results have shown that it would be difficult to
reach an international standard of quality. A EMQ less than 0,5
pixel was obtained only when images are made using new
generation cameras (RMK TOP, for instance). It must be
stressed that best results were obtained only when one or two
measurements higher than 15um have been eliminated. This is
only possible if cameras with 8 fiducial marks are used.
As can be seen 33% of measurements were higher than [5um.
Three possible sources of error can be responsible for those
results:
Intern
(i)
(ii)
(iii) Se
3.3 ]
To
aerotr
differ
Brazil
projec
from
from I
Proje
Table
adjusti
image
to be 1
during
Quality
were h
result
since |
worst 1
no goa
pixel).
parame
correct
reason
quality
Altdofi
produc
and B.
image
obtaine
Altdorf
