|
|
Mathematical modelling with L-systems
Janoutová, Růžena ; Procházková, Jana (referee) ; Martišek, Dalibor (advisor)
This work deals with L-systems and their practical utilization. The work can be devided into two parts - theoretical and applied. In the theoretical part, i.e. in the first chapter, the L-systems are discussed. There can be found their introduction and classification. Apllied part includes second and third chapter. Second chapter contains review of L-system's utilization and third chapter is dedicated to program attached to this work, his description and manual.
|
|
|
Mathematical methods of morphology modelling of coniferous trees
Janoutová, Růžena ; Štarha, Pavel (referee) ; Martišek, Dalibor (advisor)
The thesis was focused on creation of a coniferous tree by nondestructive method allowing description of structure of adult spruce trees. After processing provided data we created a model of L-system which creates a tree branch. Thereafter, a Python script generated parameters which were required for the creation of the model of the tree in graphical software Blender. Model of coniferous tree was sucesfully generated. Its memory requirements are high but for our purposes this is not an essential problem.
|
|
|
Modelling 3D Forest Structure for Improved Retrieval of Forest Biophysical Properties
Janoutová, Růžena ; Albrechtová,, Jana (referee) ; Štarha, Pavel (referee) ; Martišek, Dalibor (advisor)
Hlavním cílem práce bylo zlepšení kvantitativních odhadů vegetačních parametrů smrkových porostů pomocí spektrálních simulací trojrozměrného modelu přenosu záření. Prvně bylo potřeba vytvořit přesný 3D model smrku. Implementace přesného 3D modelu smrku pro parametrizaci celých lesních porostů je v současné době výpočetně nemožné, bylo tedy nutné tento 3D model smrku zjednodušit. Přesný 3D model smrku společně s dostupnými leteckými daty sloužil pro nalezení optimálního zjednodušení. Optimální model vedl ke kompromisu mezi výpočetní náročností a přesností výsledné odrazivosti z modelu přenosu záření. Následně byl optimální model smrku využit pro odhady vegetačních parametrů ze satelitních snímků. Přesnost odhadů byla ověřena oproti pozemním měřením odhadovaných parametrů. Na závěr byly porovnány výsledky z odhadů vegetačních parametrů pomocí optimálního 3D modelu smrku s výsledky z tradičního přístupu pomocí modelů stromu s geometricky jednodušími tvary korun.
|
|
|
Estimation of winter wheat yield using machine learning from airborne hyperspectral data
Švik, Marian ; Pikl, Miroslav ; Janoutová, Růžena ; Veselá, Barbora ; Slezák, Lukáš ; Klem, Karel ; Homolová, Lucie
Methods based on optical remote sensing allow nowadays to assess crop conditions over larger areas. The assessment of crop conditions and potential estimation of crop yields in the early growth\nstages can help farmers to better target their management practice such as application of fertilizers. In this study we analysed airborne hyperspectral images acquired several times during the growing season over two experimental sites in the Czech Republic (Ivanovice and Lukavec). The field experiments on winter wheat included 12 levels of fertilisation (combination of organic and mineral fertilisers). Such an experiment design and the possibility of combining the data from two sites together increased the variability in our wheat yield dataset, which varied between 2.8 and 10.0 t/ha. Further, we used a machine learning method – namely gaussian process regression from the ARTMO toolbox to train two variants of models: a) combining the spectral data from both sites and from the multiple acquisition days and b) combining the spectral data from both sites for individual acquisition days.The results showed that it was feasible to predict wheat yield already at the beginning of April with R2 > 0.85. This promising result, however, requires more thorough validation and therefore we plan to include more data from other sites in the next steps.
|
|
|
HYPOS aerial hyperspectral campaign
Hanuš, Jan ; Homolová, Lucie ; Janoutová, Růžena
The ESA project HYPOS (Hyper-Spectral Analysis and Heterogeneous Surface Modeling) was solved by CzechGlobe together with three French partners (Magellium, CESBIO, IRSTEA) between 2016 and 2018. The task of the project was to create algorithms for estimating vegetation parameters from hyperspectral data and to define the most suitable spectral bands for estimation of parameters from multispectral data.
|
|
|
Modelling 3D Forest Structure for Improved Retrieval of Forest Biophysical Properties
Janoutová, Růžena ; Albrechtová,, Jana (referee) ; Štarha, Pavel (referee) ; Martišek, Dalibor (advisor)
Hlavním cílem práce bylo zlepšení kvantitativních odhadů vegetačních parametrů smrkových porostů pomocí spektrálních simulací trojrozměrného modelu přenosu záření. Prvně bylo potřeba vytvořit přesný 3D model smrku. Implementace přesného 3D modelu smrku pro parametrizaci celých lesních porostů je v současné době výpočetně nemožné, bylo tedy nutné tento 3D model smrku zjednodušit. Přesný 3D model smrku společně s dostupnými leteckými daty sloužil pro nalezení optimálního zjednodušení. Optimální model vedl ke kompromisu mezi výpočetní náročností a přesností výsledné odrazivosti z modelu přenosu záření. Následně byl optimální model smrku využit pro odhady vegetačních parametrů ze satelitních snímků. Přesnost odhadů byla ověřena oproti pozemním měřením odhadovaných parametrů. Na závěr byly porovnány výsledky z odhadů vegetačních parametrů pomocí optimálního 3D modelu smrku s výsledky z tradičního přístupu pomocí modelů stromu s geometricky jednodušími tvary korun.
|
|
|
Potential of the sentinel-2 red edge spectral bands for estimation of eco-physiological plant parameters
Malenovský, Zbyněk ; Homolová, Lucie ; Janoutová, Růžena ; Landier, L. ; Gastelluetchegorry, J-P. ; Bertholt, B. ; Huck, A.
In this study we investigated importance of the spaceborne instrument Sentinel-2 red edge spectral bands and reconstructed red edge position (REP) for retrieval of the three eco-physiological plant parameters, leaf and canopy chlorophyll content and leaf area index (LAI), in case of maize agricultural fields and beech and spruce forest stands. Sentinel-2 spectral bands and REP of the investigated vegetation canopies were simulated in the Discrete Anisotropic Radiative Transfer (DART) model. Their potential for estimation of the plant parameters was assessed through training support vector regressions (SVR) and examining their P-vector matrices indicatingsignificance of each input. The trained SVR were then applied on Sentinel-2 simulated images and the acquired\nestimates were cross-compared with results from high spatial resolution airborne retrievals. Results showed that contribution of REP was significant for canopy chlorophyll content, but less significant for leaf chlorophyll content and insignificant for leaf area index estimations. However, the red edge spectral bands contributed strongly to the retrievals of all parameters, especially canopy and leaf chlorophyll content. Application of SVR on Sentinel-2 simulated images demonstrated, in general, an overestimation of leaf chlorophyll content and an underestimation of LAI when compared to the reciprocal airborne estimates. In the follow-up investigation, we will apply the trained SVR algorithms on real Sentinel-2 multispectral images acquired during vegetation seasons 2015 and 2016.
|
|
|
Mathematical methods of morphology modelling of coniferous trees
Janoutová, Růžena ; Štarha, Pavel (referee) ; Martišek, Dalibor (advisor)
The thesis was focused on creation of a coniferous tree by nondestructive method allowing description of structure of adult spruce trees. After processing provided data we created a model of L-system which creates a tree branch. Thereafter, a Python script generated parameters which were required for the creation of the model of the tree in graphical software Blender. Model of coniferous tree was sucesfully generated. Its memory requirements are high but for our purposes this is not an essential problem.
|
|
|
Mathematical modelling with L-systems
Janoutová, Růžena ; Procházková, Jana (referee) ; Martišek, Dalibor (advisor)
This work deals with L-systems and their practical utilization. The work can be devided into two parts - theoretical and applied. In the theoretical part, i.e. in the first chapter, the L-systems are discussed. There can be found their introduction and classification. Apllied part includes second and third chapter. Second chapter contains review of L-system's utilization and third chapter is dedicated to program attached to this work, his description and manual.
|
|
|
Mathematical methods of modelling the morphology of spruce trees
Janoutová, Růžena ; Novotný, Jan ; Pivovarník, Marek ; Zemek, František
Radiative transfer (RT) models are simulation tools which can be used to quantify relationships between vegetation canopy properties and observed remotely sensed data. Th is study aims at creating a spruce tree growth model as a key input for use in RT models. Th e spruce tree model is built on data obtained from terrestrial laser scanning of spruce trees. Each tree model is unique. Th is uniqueness is achieved by using L-systems which are able to simulate natural randomness while complying with the given tree parameters. L-systems are established on a theory of grammar that enables rewriting a string of symbols according to specifi ed rewriting rules. In practice, our tree models are generated in Blender visualization soft ware, implementing an algorithm written in Python. Th e algorithm generates the basic parameters of the whole tree and then creates the parameters of the spruce trunk and initial branches. Th e parameters are generated randomly within a range that is calculated from measured data. Th en each branch is grown on the basis of annual increments defi ned by fi eld measurements. Tree needles are distributed with respect to the age of individual branches; therefore, the needles have diff erent colours according to their age. Cones and faces are graphical representations of the spruce model. Branches are represented by cones and needles are represented by faces around the branches. Th e faces are transparent, thus simulating light transmittance in-between the needles. The whole model is highly computationally demanding, especially with respect to computer memory.
|
guest ::
