Merger — means merging of two neighboring objects when decreasing the scale,
e.g. parallel linear objects (contour of forest or road) in small-scale may merge. The
algorithm of lines’ simplification differs by criteria of points’ selection, namely:
1.
The algorithm of independent points is very simple and does not take into
consideration the topological bonds with other coordinate pairs: either each n-point is
rejected (e.g., each third of each tenth) or 1/n share of the set of coordinate pairs is
selected randomly.
2.
The algorithm of local processing ensures exclusion of coordinate pairs
taking into consideration adjacent points’ position, is determined by the following
criteria: Euclidean distance between the points (length of the segment of the digitized
line); change of angles between adjacent line segments; combination of the first two
criteria.
There exist different extensions of local processing algorithms, the search there is
done not only among adjacent points, but line segments are also assessed. The value of
the search is determined by different criteria, including complexity of the lines, density
of coordinate pairs, the starting point of the search.
Using the theory of fractals. Simplification and rounding, no doubt, improve the
overall appearance of the line, but take into consideration only geometry of the line
and do not pay attention to geographical peculiarities. For some objects (roads, rivers)
such methods are completely justified, but many nature peculiarities are disregarded or
simply lost.
It should be noted that some geographic objects (coasts of seas and oceans, rivers,
state boundaries etc) where they are drawn by nature contours – have fractal property
of similarity. To date, algorithms that allow to use fractal properties for generalization
of linear objects, have been developed. There is such a generalization method in which
the fractal dimension of lines is kept, e.g. with adding details of digitized or already
generalized lines at an enlarged scale.
Geospatial analysis – is a process of searching for spatial patterns in distribution
of geographical data and interconnections between objects [5]. Geospatial analysis is
a kind of unique lens used to study events, structures and processes which happen on
the earth or close to the surface of our planet. As a result of analysis of geographic
information, new information is obtained and unknown patterns are revealed.
Methods of spatial analysis operate in some spatial and time scales. When
studying spatial analysis, the following aspects are essential:
- Area. The area of geospatial analysis is the surface of the Earth, the shell over
it in the analysis of topography and atmosphere, the shell under it in the analysis of
groundwater and geology.
- Scope. The scope of geospatial analysis objects extends from small objects (e.g.
archeologists’ records about places where ceramic articles’ pieces were found,
- 278 -