22 
The instrument generates model co-ordinates (X, Y, Z) and, for orien- 
tation, the base component by. This permits a semi-automatic orientation. 
The computed values of orientation parameters are displayed and manually 
set on instrument. 
The system can be programmed for parallel profiling in any direction 
(e.g. for orthophoto printing) with adapted speed. Furthermore, the instrument 
tracking device can be positioned to any pre-determined location either in the 
model- or in terrestrial co-ordinate system. Moreover, it can be positioned to 
any location derived as a function of the. previously sampled data. 
The high level of interactivity between instrument and computer can be e e 
effectively exploited for progressive (24) or composite sampling (25) for DTM. 
In progressive sampling the computer controls the positioning of the instrument 
tracking device (X, Y, Z) and it analysis the data sampled. The process is 
initiated by sampling a coarse regular grid of points by computer support. The 
obtained height data are examined for local variability and a list of new points, 
which form a denser but incomplete grid, is compiled accordingly. The com- 
puter then steers the tracking device to all new points succesively. The fine 
settings in height are performed manually. The new and old data are then 
examined together for variability and a new list of points, to be sampled in 
the next run, is compiled. The procedure is repeated until a specified accuracy 
is achieved. Density of the resulting (incomplete) grid is locally adapted to terrain 
variability. In the vicinity of distinct morphological features (i.e. break-lines 
and break-points) the grid can be very dense or sometimes significant terrain 
features might be lost. By implementing a composite sampling technique 
these shortcomings can be prevented. 
In composite sampling the distinct morphological data are procured 
selectively under operator's control. These data can be integrated in progres- e @ 
sive sampling in different ways (25). 
Other applications exploiting the self-adapting capacity, particularly for 
engineering applications, are easy to imagine. 
2.2 Multiple systems 
Development of multiple systems with closed loops between analogue 
and digital components is at present in the conceptual stage. The basic system 
configurations are similar to those of the serial multiple systems, with the 
difference that additional links are established between digital and analogue