The workflow for a) - c) is as follows : 
1. Collection and preparation of source data. 
2. Transferring of all geometries and attributes needed for 
the DLM 25/1 from the different sources into a single 
"information sheet" which also shows how the data has to 
be structured. 
Checking of the information sheet by a supervisor. 
Data capturing by table digitizing using ALK-GIAP 
workstations. 
Running a final checking procedure. 
Running a procedure to check along the frame to 
adjacent capturing units. 
7. Cooperation with other state survey administrations for 
exchange and adaption of state boundary data. 
Conversion to the EDBS format for 
Input into the official ATKIS data base installed on an IBM 
mainframe computer. 
A O0 
eo o 
«o oo 
Most of the capturing work (steps 2 and 4) has been done in 
contract by private companies. In some cases the DLM 25/1 
was captured with the Zeiss/PHOCUS-system from stereo 
models using analytical plotters Planicomp P1/P3 with 
Videomap superimposition or a converted analog stereo- 
plotter TOPOCART. 
3. DLM 25 REVISION AND EXTENSION 
Starting in 1997 the LVermA BB is faced with the following 
tasks : 
- improvement of planimetric accuracy, 
- updating of the DLM 25/1 and 
- extension to the content of the DLM 25/2. 
The primary data sources used for establishing the DLM 25/1 
in Brandenburg did not always fulfill the planimetric accuracy 
requirement of + 3 m. 
In some cases the TK 10 N used for capturing the DLM 25/1 
was based on a revision from 1990. In some other cases the 
photo flight for producing orthophoto maps TK 10 L took 
place in 1991. For the time period up to 1997 a lot of 
changes can be expected which are partly influenced by the 
reunification of the two German states. 
The content of the second realization stage DLM 25/2 has 
already been defined by the AdV as OK-DLM 25/2. Almost all 
relevant objects of the class "settlement" with the sub-classes 
"built up areas", "open residental areas" and "buildings and 
facilities" became part of the DLM 25/2. However buildings 
were left for a further stage. For objects of other classes it 
was defined that additional attributes have to be captured. 
The LVermA BB has decided that further objects and 
attributes which are part of the complete edition of the 
OK-DLM 25 but not part of the OK-DLM 25/2 shall be 
captured or incorporated. These are e.g. trigonometric and 
levelling points, water-gauges, boundaries of different types 
of protective areas, walls, depots and buildings. 
3. 1 Alternatives 
Updating of topographical data involves the following 
processing steps : 
- detection of (or getting knowledge about) relevant 
topographical changes, 
- acquisition and examination of information about these 
changes, 
- data capturing including geometry and attributive 
informations, 
- incorporation into the data base and deletion of 
pre-existing data. 
The following alternatives had to be examined : 
a) periodical or continuous updating, 
b) homogeneous or priority treatment of certain areas or 
object classes, 
c) updating by contracts or by own staff, 
d) various methods for change detection and data 
collection. 
All users of the DLM 25 agree that a periodical revision of all 
data within a at least 5-year-cycle is absolutely necessary. 
Some users also demand a continuous updating. If changes 
can be detected or will be reported, they have to be captured 
and incorporated into the database immediately. Other users 
demand a priority treatment for densely populated areas or 
certain object classes (e.g. built-up areas and roads) with a 
shortened revision cylce of one year. 
These additional demands can not be fulfilled with 
photogrammetric methods since too many photo flights 
would be necessary and the capacity for orthophoto 
production is limited. A close contact and data exchange 
between the LVermA BB and administrations, other 
institutions and private industry which can report changes 
and supply digital data could solve the problem in the 
long-term. At the moment almost all of them are still in a 
stage of building-up their information systems. 
An active terrestrial reconnaissance service, as proposed by 
the state survey administration of Northrhine-Westfalia, would 
require personnel capacities which are not available at the 
LVermA BB. By various reasons the cadastral survey offices 
can not contribute to this work. Furthermore a continuous 
updating and/or a priority treatment would cause an 
increased effort for data management. 
3.2 General Concept 
Possible updating methods and workflows have been 
investigate and weighted. For the finally choosen procedures/ 
solutions the specific conditions in Brandenburg had to be 
taken into consideration. At the present stage of evaluation 
the following strategy seems to be most efficient at minimum 
costs : 
- Improvement of planimetric accuracy should be done as 
a first independent step. 
- Updating of the DLM 25/1 and extension to DLM 25/2 
should be done as a common step in parallel to the 
5-year-cycles for 
a) revision of the TK 10 N and 
b) digital generation of new orthophoto maps (TK 10 
DL). 
The hang on to these two cycles offers the following 
benefits : 
- The work flow for updating the TK 10 N and the DLM 
25/1 has some steps which are very similar. By 
combining both tasks double work can be avoided. 
- The cycle for digital generation of the TK 10 DL follows 
the cycle for updating the TK 10 N with a delay of 2 to 3 
years. A hang on to both cycles guarantees an 
up-to-dateness of 3 years for the DLM 25. This would be 
a good compromise between periodical homogenious 
updating and continuous updating or priority treatment of 
areas or object classes. 
582 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B4. Vienna 1996