|
|
Evalution of Time Series of Satellite Images to Observe Disturbancec in the Low Tatras
Laštovička, Josef ; Štych, Přemysl (advisor) ; Hais, Martin (referee)
The work is aimed at finding appropriate methods for observing changes in the status of forest vegetation and its evaluation in the years 1992-2015. The satellite images of the Low Tatras are analyzed by using Time Series technology. Specifically, the images Landsat 4, 5, 7 and 8, for which it is necessary to perform a calibration and an adjustment of input data values to realize the individual vegetation indices, due to the fact that the images are captured by different sensors with different radiometric resolution. From this perspective, the work deals with the possibilities of normalized relative radiometric corrections and search for a particular type of appropriate compensation for Landsat CDR images. Calibrated data sets are evaluated by Time Series of different vegetation indices. The resulting values are evaluated in relation with the occurrence of forest disturbances, eg. wind storms, biological and other pests. The final part is discussion of the results, evaluating the test methods of calibration and suitability of vegetation indices for observing the state of calamity. The App is created for generating the Time Series of Landsat images CDR and for preparing RRN datasets. Key words: Time Series, radiometric correction, atmospheric correction, Landsat CDR, vegetation indices,...
|
|
|
Land cover changes in military areas of Czechia
Outrata, David ; Štych, Přemysl (advisor) ; Červená, Lucie (referee)
1 Land cover changes in military areas of Czechia The aim of this bachelor thesis was to analyze and compare land cover changes in two areas affected by military activities, in military area Brdy and former military area Ralsko between 1986, 1998 and 2011. Maximum Likelihood supervised classification algorithm and Landsat 5 satellite images was used. Within the work was also examined the usability of Landsat images CDR, with atmospheric corrections applied. Reference data to gain control data were RGB and monochromatic aerial images. The classification system contained eight classes adapted to explored territory. Satellite photographs was classified on surveyed territories for the years 1986, 1998 and 2011, with overall classification accuracy ranged from 80.45% to 90.02%.With these data was further worked, and tables and graphs showing the area of individual land cover in the given time horizons was compiled. Outputs are also land cover maps, change maps and stable land cover maps. The expected trend that in the former military area after leaving the army land cover changed significantly, contrary to the current military area, where changes are minor, was confirmed. Key words: land cover changes, supervised classification, Landsat, Landsat CDR, military areas, Brdy, Ralsko, Geomatica, ArcGIS
|
guest ::
