Remote control technologies on land cover elements classification and agrarian resources assessment



V. I. Voloshyn and V. M. Korchinsky

Dnipropetrovsk, Ukraine


 
Remote sensing of earth using multispectral observation platforms has now become an important part of a number of diverse human activities, such as
cartography, monitoring of environment, assessment of security schemes, etc.  The image databases that we use to validate our studies are obtained in the EO
range of the electromagnetic spectrum with such satellites as: Sich-1M, Meteor - 3M, IRS (LISS), Spot, Landsat, Terra (Aster), etc. However, the present
methods are more extensively studied with Ukrainian users satellites image databases such as Sich-1M and Meteor -3M.

A method for radiometric correction of multispectral images to ensure enhancement of information content and spatial resolution was developed. The method
offers correction in the case of point displacement between different spectral channels due to diffraction effects and spatial coordinate-susceptibility of photo
receivers causing positioning instability during sensing.


As applications of this method, some case studies on land cover elements classification and agrarian resources assessment are worked out. In particular the
complex biophysical measurements including the vegetation structure like NDVI was assessed for some areas of representing various landscapes, geographic
and climate zones. Formal sites are based on high-resolution remote sensing, ancillary and field plot data.


The development of the test site database for selected Ukrainian territories is intended to represent the earth's diversity of land covers and types of land cover change. The site network represents global, regional and local change processes due to both natural and anthropogenic factors. Test criteria include phenological class (seasonal grassland, deciduous forest), anthropogenic (urbanization, agriculture, conversion, biomass burning), interface (land/snow, land/water), biotic
(insect and pathogen), and hydrologic (seasonal inundation) representation. A number of critical sites and "hotspots" sites is included in the network because of their particular conservation, political, economic and social significance. As applications of this method, some case studies on potentially dangerous landslide areas selection and classification of land cover elements classification have been developed as well.