286 
  
  
  
Fig. 1: UMK-SCAN macro-scanning stereo system Fig. 2: JenScan micro-scanning stereo system 
Sensor Scanning Principle Max. Image Size Pixel Size Calibr. Focal Length 
Lens 
UMK-SCAN macro-scan 15414 x 11040 pixels 11x 11 um 100 mm 
Lamegon 8/100 B 
  
JenScan micro-scan 4608 x 3072 pixels 11 x 11 um 14 mm 
Lamegon 3.5/14 
  
Kodak DCS 420 mi still video 1524 x 1012 pixels 9 x 9 um 14 mm 
> Nikkor 3.5/18 
  
JenaCam fixed chip 768 x 512 Pixel 11 x 11 um 14 mm 
Lamegon 3.5/14 
  
  
  
  
  
  
  
Table. 1: InduSURF Digital - sensors and lenses used 
2.2 Digital image orientation: DIAMANT point measuring software and BINGO bundle triangulation program 
For image orientation, the control points and model tie points are defined by retroreflecting target marks. The 
measurement of these points in the digital images is performed using the DIAMANT point measuring software 
developed by Carl Zeiss in Jena. This software includes interactive and automatic measurement algorithms for locating 
and measuring the retro-reflecting targets. For the exact definition of the point centers, digital measurement operators 
have been developed, which work more objectively and precisely than the human eye and display a high degree of 
resistance to irregularities in the point geometries. As an alternative, however, any target in the image can also be 
measured by using different cursor shapes and sizes and with the aid of zoom functions. 
The following measurment operators have been implemented in DIAMANT: 
OQ) Centroid operator: edge detection using threshold value, point center by weighted averaging 
O Ellipse operator 1: edge detection using Laplace filter, point center by adjustment of the edge points 
O Ellipse operator 2: edge detection using moment preservation, point center by adjustment of the edge points 
O Mouse operator: zooming is possible, point center by positioning of the mouse pointer on the monitor 
After measurement of the image, the pixel coordinates obtained and all other project data are edited for the BINGO 
bundle adjustment program (Technical University of Hanover, GIP Aalen), and the BINGO input files are automatically 
generated. BINGO is then used to define the elements for the exterior orientation of the digital images and to read out 
the accuracies for the measured and computed image or object coordinates. 
The new BINGO software release also permits the computation of free network adjustment. In this process, quantities 
such as X, Y, Z , Phi, Omega, Kappa and scale can be grouped as freely selectable parameters into meaningful 
combinations. As a result, it is now possible to adjust the image block without any geodetic control data being required. 
In practical applications, however, one frequent requirement is the transformation into a reference coordinate system 
specified by the user, which means that reference points or other object data are already known or need to be 
measured into the system beforehand. 
IASPRS, Vol. 30, Part 5W1, ISPRS Intercommission Workshop “From Pixels to Sequences“, Zurich, March 22-24 1995