|
|
Deep learning for tree line ecotone mapping from remote sensing data
Dvořák, Jakub ; Potůčková, Markéta (advisor) ; Lefèvre, Sébastien (referee)
Deep learning is growing in popularity in the remote sensing community, especially as a classification algorithm. First part of this thesis describes deep neural networks commonly used for remote sensing classification and their various applications. Capabilities of selected geospatial software suites in relation to deep models are also discussed in this part. Theoretical findings from the first part of the thesis are validated using two deep convolutional Encoder-Decoder networks - U-Net and its proposed adaptation called KrakonosNet. They are used to perform a sematic segmentation of spruce trees and dwarf pine shrubs in the tree line ecotone of the Krkonoše Mountains, Czechia. A normalised digital surface model is employed for creation of sufficiently large amount of training data, while the classification itself is performed using only optical imagery with very high spatial resolution. Resulting classification is compared to a set of traditional remote sensing classifiers, namely Maximum Likelihood, Random Forest, and a Support Vector Machine. Both U-Net and KrakonosNet significantly outperform the other classifiers on this dataset and will be consequently used in a related research project. Key words deep learning, U-Net, Krkonoše mountains, classification, vegetation mapping, picea abies,...
|
|
|
Blocky accumulation mapping from RPAS LiDAR and image data
Kolář, Michal ; Potůčková, Markéta (advisor) ; Dušánek, Petr (referee)
With merit of constant development in measuring technology it is possible to obtain data of high resolution and accuracy describing Earth's surface. During the project "Vegetation and Krkonoše tundra change detection method development by analyzing data from multispectral, hyperspectral and LiDAR UAV sensors" high quality data were acquired, with point density reaching up to 800 points/m2 and orthophoto of GSD 0.02 m. Data are capturing cryoplanation terraces in NE parth of Luční hora in three time periods: June, July and August 2019. The aim of this work is to devise a methodology of blocky accumulation mapping and evaluating detail of data. Key words: blocky accumulation, laser scanning, UAV, point cloud, orthophoto, segmentation
|
guest ::
