- Supervised classification. The algorithm classifies by the mini- 
mum distance decision rule per point, For distance measure Eucle- 
dian'is used./Fig.3/ 
All programs are written in Fortran IV language. Efforts are under | 
way to make further progress in texture analysis, automatic segmen- 
tation of homogeneous areas, hierarchial classification using the 
optimal combination of spectral bands as well as upgrading system 
functions like menu extension in conversational mode and so on. 
THE PUBLIC RELATIONS 
There are some promotional activities provided by the Remote Sen- 
sing Department like exhibitions /e.g. Budapest Int'1 Spring 
Fair'82/ presentations /twice a month in general/ and so called 
open days with hands-up option for present and future users. In 
addition, the staff members are usually requested to give lectu- 
res on remote sensing, multispectral data analysis etc. at univer- 
sities, postgraduate schools and scientific organisations. Leaflets, 
technical newsletters help to share information with over 80 poten- 
tial users. Survey sheets ensure feed-back /comments, suggestions/ 
from end users. 
As far ‘as the publication activity is concerned, some papers have 
been published in the "Geodézia és Kartográfia" journal issued 
bi-monthly in Budapest. 
  
RESEARCH AND DEVELOPMENT 
In the last two years several projects have been carried out in 
various application fields like regional planning, water quality 
mapping and land use. Research teams work on the development and 
implementation of analogue and computer-aided interpretation 
methods in close cooperation with other institutions and agricul- 
tural companies. 
Some of these works are based on use of test areas. One of them is 
the multipurpose Kiskóre Test Field at the newly constructed dam 
on the river Tisza /Fig.4/ The aim of the investigation is to ob- 
serve, recognize, classify and identify land use, forest and water 
quality categories using multitemporal Landsat and six wavelenght 
band multispectral imageries with some in situ measurements. In 
case of crops, the visual interpretation combined with digital 
analysis has led to the identification of the several categories 
with a degree of accuracy over 80%. The ground truth data were 
traditional land use maps with some description abouth growth 
stage, the agronomic technology applied, genetic soil maps as well 
as agrometeorologic observations. 
An another test field is located near to Debrecen. It serves as ex- 
perimental test area especially for agricultural applications/land 
use mapping, mapping of soils, vegetation analysis, damage detec- 
tion, biomass and quality estimation for crops/. The work is 
started some months ago. The main goal is to investigate the possi- 
bilities, how the remote sensing technology can operational improve 
the decision-making related to the renewable natural resources li- 
ke forests, soils and vegetations. Additional important and a 
succesful experiment has been carried out in November last year, 
  
In cooperation with the Polish Remote Sensing Centre OPOLIS, an Fi 
airborne thermal profiler /AGA THP-1/ mission was executed in Hun- te 
gary, mapping heat losses in urban environment, monitoring heat we 
polutions and analysing cooler pond’s effectivness as well as fo 
executing vegetational classification. 
100 
3E; "y d