= 20 + 
changes of reference systems can be affected on the plotting table. 
The results of this type of orientation may be understood as the 
unfolding of a polyhedron into a plane. The same applies to the possi- 
bility of unfolding curved surfaces. The compilation program may be 
arranged so that the handwheels move the measuring mark in an ortho- 
gonal curvilinear system of coordinates, and the footwheel moves the 
mark along the normals to the reference surface. It is evident that 
the same projection techniques may be applied to the sets of coordinates 
that are to be stored in digital form or presented in graphical form on 
plotting tables or CRT displays [19], [20], [21]. 
After this short description of several procedures for solving 
problems in close-range photogrammetry, it is not difficult to assess 
the special role that analytical instruments should play in this branch 
of photogrammetry. It is also easy to realize that the limitations of 
classical photogrammetric instrumentation must be one of the main causes 
for the slow development of close-range photogrammetry, in spite of the 
high interest in the application of photogrammetry to many branches of 
science, engineering and medicine. 
6. CONCLUDING REMARKS 
  
Having reached the point in the development of analytical 
instruments where the attainable universality permits an unprecedented 
flexibility in the software design with considerable ease of programming, 
a concentrated effort should be directed towards the theoretical and, 
above all, experimental assessment of the potential advantages that have 
been discussed in this paper. Three main classes of problems should be 
considered. 
The first one concerns the problems related to the inclusion 
of analytical instruments into the larger information processing 
systems. In general, an operation performed on an analytical instrument 
is seldom an end in itself. When a system approach is taken to the 
solution of a practical task the operation performed on an analytical 
instrument is definitely a component of a broader information gathering 
and processing scheme. Except for the clear cut cases where no other 
instrument can perform the task the main questions to be answered are: 
Does the inclusion of an analytical instrument significantly improve 
the performance of a particular system? How does the choice of a parti- 
cular procedure on an analytical instrument influence the performance of 
the information system as a whole, and what are the cost-to-benefit 
ratios for different choices of procedures under particular circum- 
stances (e.g. operation in a digital cartographic system, or a system 
for production of orthophotos)? 
The second class of problems that follows from the first one 
is centered around the question of the level of universality. It is 
quite obvious that institutions active in research and development, and 
firms that are involved in photogrammetric activities of considerable 
diversity will be best served by high precision universal analytical 
instruments. The case is somewhat different with organizations that 
specialize in a single or a small number of similar applications with a 
large volume of work extended over long periods of time. Even if they 
need the facility for on-line analytical processing with feedback, 
because of some requirements that cannot be fulfilled at all, or not as