209
geointormation structures.
Mathematic model of composition. On the basis of the
presented theoretical propositions a mathematic model of
composition for analysis and differentiation of
polymorphic-isomerical structures of geoinformation has been
worked out. . In this model probability estimation of
interdependency between components within its structure is
used. Consider the structure of the model. Let each point
(pixel) of the remotely sensed object, of be represented with
the set X- ' L = 1 , 2, ,k> of values of intensity of
refi ec t: ed rad iati on 'si goal s >
each point of the
Ject
>et of compositions of various signals
f
Z J *
k a > is realised. As a measure of relationship between values
of intensities in composition os will use probability of
their joint realisatierc To describe probability structure
c.*t
composi t i. on s.
1 et.'s
appi y
family óf
a ws o -
: stribution of probabilities C Pj (kj)1 , j = 1
intensities, which gives a full description
structure of composit ions of 1 oformation field
the composition model of the image the values of
in composilions are fixed by means of intervals
the type of intordependency between them.
of i nn&r
(image). In
intensities
given, and
wh iCh i S
d e t er m i n ed according to
t=j
P- (kj ) n P. (X; >, j=2, 3, ...,n(l>,
J J L =2 L
may change. Therefore, the difference between information
structures of relations between the components it is
composed of, and which is coded by the value of the type of
interdependency between them in accordance with (1). Three
situations are estimated which mean correspondence,
norrcorrespondence arid independence between components, and
are coded with signs +, 0. The composition code is
constructed as follows. All the k of intensities are
examined and the value of the criterion is estimated, then
are sequentially eyainined combinations Ck - 1), (k — 2) ,
, e ,, 2 of intensities, and the values of the criterion are