|
|
Updating of land use cadastral records using satellite data Sentinel-2
Šaffová, Michaela ; Štych, Přemysl (advisor) ; Kupková, Lucie (referee)
Updating of land use cadastral records using satellite data Sentinel-2 Abstract The aim of the project is to propose a methodical procedure which will classify selected land use classes with an accuracy of more than 80 %. Constructed methodical procedure will have the task of detecting the discrepancy of land reported in the cadastral records with the real state in the landscape identified by satellite Sentinel-2 data classification. Classification of agricultural land classes (arable land, permanent grassland, orchard and vineyards) is solved in this project using multitemporal data Sentinel-2 using object classification methods. The first part of thesis focuses on the literary introduction to the topic of the theme. The second part is devoted to the process of creating a methodical procedure for object classification of land use classes, where parameters are defined by experimental activity and thresholds of the defined classification. The results of the work are compared and evaluated using the overall accuracy and error matrices of the classification using the developed algorithm. Keywords: cadastral records, multitemporal satellite imagery, object-base classification, Sentinel-2, agricultural land resources
|
|
|
Classification of selected species of vegetation in the Krkonoše Mountains tundra based on time series of PlanetScope imagery
Roubalová, Markéta ; Kupková, Lucie (advisor) ; Červená, Lucie (referee)
The aim of this thesis was to test the suitability of PlanetScope imagery to differentiate and evaluate the possibility of multi-temporal approach to improve classification accuracy of selected vegetation species (Molinia caerulea, Calamagrostis villosa, Nardus stricta) in eastern tundra in the Krkonoše Mts. National Park. PlanetScope imagery - 4 spectral bands with spatial resolution 3 m - was used. Per-pixel classifications Maximum Likelihood, Support Vector Machine and Random Forest and object-based classification SVM were executed in software ENVI 5.3. based on GPS field data collected from 2014 till 2018. The best classification results were compared to classification results in Kupková et al. 2017 and Marcinkowska-Ochytra et al. (2018a). The overall accuracy of the best classification result (multitemporal composite using Random Forest classifier) was 80,67 %. It is better result than in the case of single image classification (overall accuracy was 76,06 %). PlanetScope data were compared to RapidEye and Apex data. The overall accuracy of the RapidEye best classification result (multitemporal composite using Random Forest classifier) was 74,75 %, the best overall accuracy of monotemporal classification of Apex data reached 75,91 %. Key words: multi-temporal classification, vegetation,...
|
|
|
Training and validation dataset optimization for Earth observation classification accuracy improvement
Potočná, Barbora ; Kupková, Lucie (advisor) ; Potůčková, Markéta (referee)
This thesis deals with training dataset and validation dataset for Earth observation classification accuracy improvement. Experiments with training data and validation data for two classification algorithms (Maximum Likelihood - MLC and Support Vector Machine - SVM) are carried out from the forest-meadow landscape located in the foothill of the Giant Mountains (Podkrkonoší). The thesis is base on the assumption that 1/3 of training data and 2/3 of validation data is an ideal ratio to achieve maximal classification accuracy (Foody, 2009). Another hypothesis was that in a case of SVM classification, a lower number of training point is required to achieve the same or similar accuracy of classification, as in the case of the MLC algorithm (Foody, 2004). The main goal of the thesis was to test the influence of proportion / amount of training and validation data on the classification accuracy of Sentinel - 2A multispectral data using the MLC algorithm. The highest overal accuracy using the MLC classification algorithm was achieved for 375 training and 625 validation points. The overal accuracy for this ratio was 72,88 %. The theory of Foody (2009) that 1/3 of training data and 2/3 of validation data is an ideal ratio to achieve the highest classification accuracy, was confirmed by the overal accuracy and...
|
|
|
Lost landscapes of mountain agriculture of Šumava (Bohemian Forest): reconstruction, 3D models of landscape and possibilities of their presentation on web
Polák, Mojmír ; Kupková, Lucie (advisor) ; Štych, Přemysl (referee)
The study deals with the model area of the village Prášily in Šumava (Bohemian Forest), which is analysed in the framework of the project "Heritage of lost landscapes: identification, reconstruction and presentation". The goal was to evaluate landscape changes since the middle of the 19th century until present with use of maps of stable cadaster, historical orthophotos, current ortophotos and to create 3D reconstruction models of the landscape and a 3D photorealistic model of the extinct village Hurka. The next aim was to evaluate possibilities of presentation of 3D models on the internet and to find an optimal solution for their web presentation. The theoretical part describes possible methods of evaluation of landscape transformation, development of landscape in the Czech Republic since the middle of 19th century, theory of 3D models, the use of 3D models in practice and the methods of presentation of 3D models. The practical part consists of two landscape models, that were visualized using ArcGIS Online Web Scene, Web Viewer and a 3D model of the extinct village Hůrka with fly- through animation. The models were created using Sketchfab, 3DWarehouse and Google Poly. Sketchfab was chosen as the best solution because of its good visuals and good display on mobile devices.
|
|
|
Identification of potentially suitable habitats for occurrence of European Ground Squirrel (Spermophilus citellus) using remote sensing
Kadeřábková, Tereza ; Kupková, Lucie (advisor) ; Hais, Martin (referee)
The aim of the thesis was to evaluate the possibilities of remote sensing (RS) data with different spatial resolution (UAV data with 5cm resolution, RapidEye satellite data with 5m resolution and Sentinel-2A data with 10m resolution) and of remote sensing methods (unsupervised and supervised classifications, vegetation indices NDVI and TVI) to identify potentially suitable habitats for European ground squirrel. The analyses were first carried out in Velké Pavlovice small area of interest and consequently in a broader area of interest comprised of five regions of Moravia and Slovakia. Data form mapping of biotopes and squirrels' burrows collected within a project "Sysli pro krajinu, krajina pro sysly" (European ground squirrels for landscape, landscape for European ground squirrels) were also used for the analyses. Remote sensing methods were first tested in Velké Pavlovice area using RapidEye data. The method providing the best results in the detection of European ground squirrel burrows was then used for a burrows detection in the broader area of interest (five regions) using Sentinel-2A data. The accuracy of results was defined as a proportion of the burrows detected by the resulting layer derived from RS data to the overall number of burrows mapped in the field. Best results were obtained for...
|
|
|
Comparison of changes of land cover and landscape structure in Czech and Austrian borderland in the period 1991 - 2016 using remote sensing
Brůžek, Jindřich ; Kupková, Lucie (advisor) ; Štych, Přemysl (referee)
The aim of the thesis was to compare the development of the land cover and the land use structure in the border areas of the South Moravian Region and Lower Austria on the Czech and Austrian border with the use of data and methods of remote sensing (using the method of Maximum Likelihood). Data from Landsat 5, Landsat 8 (1991 and 2016) with a spatial resolution of 30 m were used for wider area and data from the Rapid Eye with a 6.5 m spatial resolution (2010 and 2016) were used for smaller area of interest Valticko region. The classification legend was derived from the work of Rašín and Chromy (2010). The land cover classification was conducted in the ENVI 5.4 software using training and validation datasetsacquired from the orthophotos and the LPIS database. The best result of the Rapid Eye data classification in 2016 reached overall accuracy 94.87% and the worst reset was reached for Landsat 5 data (88.71%). As for the intensity of land cover changes the most interesting result was obtained for the category arable land, which on the Czech side lost 5.67 percentage points in 2016, while on the Austrian side it was a loss of only 2.81 percentage points. Overall, there were larger changes for data from Landsat satellites, which compared a longer time period. Lower changes have been found out for...
|
|
|
Classification of selected agricultural crops from time series of Sentinel-2 and PlanetScope imagery in Kutnohorsko model area
Kuthan, Tomáš ; Kupková, Lucie (advisor) ; Potůčková, Markéta (referee)
Classification of selected agricultural crops from time series of Sentinel-2 and PlanetScope imagery in Kutnohorsko model area Abstract The thesis is focused on the analysis of spectral characteristics of selected agricultural crops druring agriculutural season from time series of Sentinel -2 (A and B) and PlanetScope sensors in the model area situated around the settlements of Kolín and Kutná Hora. It is based on the assumption that the use of multiple dates of image data acquired crops in different phenological phases of the crops allows better identification of crop species (Lu et al., 2004). The aim of the thesis was to analyse the characteristics of the seasonal course of spectral features of selected agricultural crops (sugar beet, spring barley, winter barley, maize, spring wheat, winter wheat, winter rape) and to determine the period of the year suitable for the differentiation of individual crops. Another aim of the thesis was to classify these crops in the model area from time series of two above-mentioned sensors and to compare the accuracy of the pixel and object-oriented classification approach for multitemporal composites and the accuracy for monotemporal image from the term when the individual crops are clearly distinguishable. The training and validation datasets and the classification mask...
|
|
|
Laboratory/Field Spectroscopy and Remote Sensing Image Data for Vegetation Studies
Červená, Lucie ; Kupková, Lucie (advisor) ; Pavelka, Karel (referee) ; Müllerová, Jana (referee)
Dominant vegetation species of two structurally and functionally different montane ecosystems were studied by means of laboratory and field spectroscopy and remote sensing image data: (1) a homogeneous human-influenced evergreen coniferous forest represented by a Norway spruce forest in the Krušné hory Mountains and (2) a heterogeneous natural ecosystem of a relict arctic-alpine tundra in the Krkonoše Mountains with predominance of grasses. The first part dealing with the Norway spruce forest is especially focused on the methods of laboratory spectroscopy. The assessment of Norway spruce stands on a regional and a global scales requires detailed knowledge of their spectral properties at the level of needles and shoots in the beginning, but ground research is very time-demanding. Open spectral libraries could help to get more ground-truth data for subsequent analysis of tree species in forests ecosystems. However, the problem may arise with the comparability of spectra taken by different devices. The present thesis focuses on a comparability of spectra measured by a field spectroradiometer coupled with plant contact probe and/or two integrating spheres (Paper 3) and proves the significant differences in spruce needle spectra measured by the contact probe and integrating sphere, spectra of...
|
|
|
Tree species classification using sentinel-2 and Landsat 8 data
Havelka, Ondřej ; Štych, Přemysl (advisor) ; Kupková, Lucie (referee)
The main objectives of this master thesis are to evaluate and compare chosen classification algorithm for the tree species classification. With usage of satellite imagery Sentinel-2 and Landsat 8 is examined whether the better spatial resolution affects the quality of the resulted classification. According to past case studies and literature was chosen supervised algorithms Support Vector Machine, Neural Network and Maximum Likelihood. To achieve the best possible results of classification is necessary to find a suitable choice of parameters and rules. Based on literate was applied different settings which were subsequently evaluated by cross validation. All results are accompanied by tables, charts and maps which comprehensively and clearly summarize the answers to the main objectives of the thesis.
|
|
|
Mapping of terrain relics of the extinct village of Palohlavy (Ralsko) using early maps and airborne laserscanning data
Turek, Matěj ; Lysák, Jakub (advisor) ; Kupková, Lucie (referee)
Bachelor thesis deals with mapping of extinct villages using old maps, orthophotos and ALS. Specifically, it focuses on the village of Palohlavy in the former MTA Ralsko. At this location, the methodology of mapping extinct villages in this military area was tested. In the course of the work, the collection and evaluation of available data sources, field measurements, proposal and fulfillment of digidal spatial database and map creation, took place. The spatial database contains elements to determine their time existence.
|
guest ::
