Yuri Knizhnikov 
  
The produced discrete images were obtained by scanning of initial space photographs on a complex of INTERGRAPH 
and by using of the software package PHOTOSHOP. The initial image of a settlement (small city) is obtained by KVR- 
1000 and is increased up to a scale 1:10 000 (resolving power of the increased image — 2 mm ). 
The parameters of scanning were established for obtaining the information of the first geoinformation level with 
allowance for of geometrical average sizes of constitutions in the given settlement (10 x 15 m), showed on the 
topographic map of a scale 1:10 000, and also with allowance for of representational properties of an initial image. 
According to the formula (2) sizes of a pixel of a produced discrete image have compounded 0,18 x 0,18 mm (figure 4 
— the first geoinformation level), that provides the spatial resolution of a discrete image of 2,5 m. According to 
calculated pixel size the initial image was digitized under aperture of 0,18 mm and of 0,18 mm. The value of the 
modulation transfer function (MTF) for the images of interpreted objects of the given geoinformation level at optimum 
parameters has compounded 0,97 (formula 6). The decrease of photographic contrasts has not exceeded 3 %. 
The digitizing parameters were calculated for extraction of the information of the second geoinformation level with 
allowance for of geometrical mean sizes of large buildings (20 x 70 m) describing a plan of habitation quarters of 
settlement. The sizes of the pixel of produced discrete image have compounded 0,72 x 0,72 mm (figure 4 — the second 
geoinformation level), that provides the spatial resolution of a discrete image of 10 m. Further by programmatic way 
the obtained image was adduced to a scale of a updated map (1:25 000), thus the obtained image was visually accepted 
as continuous. The value of MTF for the images of interpreted objects of the given geoinformation level at optimum 
parameters has compounded 0,90. The decrease of photographic contrasts has not exceeded 10 96. 
The scanning parameters were established for obtaining the information of the third geoinformation level with 
allowance for of mean sizes of habitation quarters in settlement (100 x 150 m). The sizes of the pixel of produced 
discrete image have compounded 1,8 x 1,8 mm (figure 4 — the third geoinformation level), that provides the spatial 
resolution of a discrete image of 25 m. The obtained image was adduced to a scale of an updated map (1:50 000), thus 
the obtained image was visually accepted as continuous. The value of MTF for the images of interpreted objects of the 
given geoinformation level at optimum parameters has compounded 0,97. The decrease of photographic contrasts has 
not exceeded 3 %. 
The analysis of the results of discrete images interpretation, introduced in figure 4, has shown, that they respond the 
norms of the topographic mapping of the defined scales. The optimum level of a generalization of produced discrete 
space images is determined on the basis of obtained relationship between sizes of a pixel and geometrical mean size of 
interpreted objects (section 3), and also by calculation of representational and photometric parameters of produced 
image (formulas 2 and 6). 
5 SPECIFIC PROPERTIES OF STEREOSCOPIC VISUAL OBSERVATION OF DISCRETE AEROSPACE 
IMAGES 
The qualitative information on the spatial form of interpreted objects is traditionally received on a stereoscopic pair of 
photographic (analogue) images. It is known that the person accepts the depth of three-dimensional space due to a lot of 
features, which ones are founded on activity of physiological and psychological mechanisms of a visual system. The 
greatest value for stereophotogrammetry and interpretation have the kinesthetic mechanism of a convergence indicated 
by D.Bercly in 1709, and mechanism of binocular disparity established by C. Witston in 1839, which one has opened an 
artificial stereoscopic effect and has invented the stereoscope. 
5.1 Geometrical fundamentals of forming of a stereoscopic image 
In modern psychology of visual perception there are two basic approaches to explanation of a phenomenon of 
stereoscopic vision. One of them esteems a plastic of a spatial image of observed object as reacting to stimulant, which 
one is the binocular parallax; at the other approach it is considered, that the person directly accepts the information on 
relative arrangement of observed points. It is known that the first approach is adopted in a photogrammetry. 
Binocular parallax Ay;=yy-yc of two different points of space (C and M) and range difference up to these points AL 
are connected by following base relation of a stereoscopy: 
  
508 International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part B3. Amsterdam 2000. 
  
In € 
ster 
The 
dete 
per 
ster 
Fro 
The 
ster 
5.2 
It is 
cor 
resi 
p ix, 
Ho 
nat 
ste 
det 
for