A5019694508- Remote Sensing in Agriculture
- Diverse Scientific Research in Ukraine
- Land Use and Ecosystem Services
- Environmental Sustainability and Technology
- Agriculture and Biological Studies
- Scientific Research and Studies
- Distributed and Parallel Computing Systems
- Land Use and Management
- Soil and Environmental Studies
- Remote-Sensing Image Classification
- Remote Sensing and Land Use
- Agriculture Market Analysis Ukraine
- Smart Agriculture and AI
- Advanced Data Processing Techniques
- Environmental and Biological Research in Conflict Zones
- Remote Sensing and LiDAR Applications
- Advanced Computational Techniques and Applications
- Flood Risk Assessment and Management
- Fire effects on ecosystems
- Scientific Research Methodologies and Applications
- Big Data Technologies and Applications
- Marine and environmental studies
- Climate change impacts on agriculture
- Geographic Information Systems Studies
- Scientific Computing and Data Management
National Technical University of Ukraine “Igor Sikorsky Kyiv Polytechnic Institute”
2016-2025
Space Research Institute
2015-2024
State Space Agency of Ukraine
2008-2023
National Academy of Sciences of Ukraine
2008-2023
Institute of Physics
2014-2022
Intel (United States)
2022
Computer Algorithms for Medicine
2021
National University of Life and Environmental Sciences of Ukraine
1999-2018
Space Research Institute
2005-2009
Deep learning (DL) is a powerful state-of-the-art technique for image processing including remote sensing (RS) images. This letter describes multilevel DL architecture that targets land cover and crop type classification from multitemporal multisource satellite imagery. The pillars of the are unsupervised neural network (NN) used optical imagery segmentation missing data restoration due to clouds shadows, an ensemble supervised NNs. As basic NN architecture, we use traditional fully...
Many applied problems arising in agricultural monitoring and food security require reliable crop maps at national or global scale. Large scale mapping requires processing management of large amount heterogeneous satellite imagery acquired by various sensors that consequently leads to a “Big Data” problem. The main objective this study is explore efficiency using the Google Earth Engine (GEE) platform when classifying multi-temporal with potential apply for larger (e.g. country level)...
The convergence of new EO data flows, methodological developments and cloud computing infrastructure calls for a paradigm shift in operational agriculture monitoring. Copernicus Sentinel-2 mission providing systematic 5-day revisit cycle free access opens completely avenue near real-time crop specific monitoring at parcel level over large countries. This research investigated the feasibility to propose methods develop an open source system able generate, national scale, cloud-free...
For many applied problems in agricultural monitoring and food security, it is important to provide reliable crop classification maps. Satellite imagery extremely valuable source of data maps a timely way at moderate high spatial resolution. Information on parcel boundaries that takes into account the context may improve quality compared pixel-based approaches. In general, parcels contain several plots with different crops such situations should be taken when using boundaries. this paper, we...
Ukraine is one of the most developed agricultural countries in world. For many applications, it extremely important to provide reliable crop maps taking into account diversity cropping systems used Ukraine. The use optical imagery only limited due cloud cover, and previous studies showed particular difficulties discriminating summer crops such as maize, soybeans, sunflower, sugar beet. This paper focuses on exploring feasibility assessing efficiency using multitemporal satellite...
In this article, the use of time series satellite imagery to flood hazard mapping and risk assessment is presented. Flooded areas are extracted from images for flood‐prone territory, a maximum extent image each event produced. These maps further fused determine relative frequency inundation (RFI). The study shows that RFI values water depth exhibit same probabilistic distribution, which confirmed by Kolmogorov‐Smirnov test. produced map can be used as map, especially in cases when modeling...
This paper focuses on drought risk assessment using satellite data. Methods of the extreme value theory (EVT) are applied for a time series vegetation health index (VHI) derived from National Oceanic and Atmospheric Administration satellites in order to provide hazard mapping. A Poisson-GP (generalized Pareto) model is modelling VHI values. The allows estimation mapping return periods different categories droughts. An approach economical due droughts presented that relies following...
For accurate crop classification, it is necessary to use time-series of high-resolution satellite data better discriminate among certain types. This task brings the following challenges: a large amount for download, Big processing and computational resources utilization state-of-the-art classification approaches. solving these problems, we have developed an automated workflow, which based on machine-learning techniques. By deployment workflow cloud platform, can overcome challenges...
Along the season crop classification maps based on satellite data is a challenging task for countries with large diversity of agricultural crops different phenology (crop calendars). In this paper, we investigate feasibility delivering early and along specific using available free over multiple years, including Landsat-8, Sentinel-1 Sentinel-2. For study, test site in Kyiv region (Ukraine) selected, which have been collecting ground types every year since 2011. Crop type are generated...
Abstract. One of the problems in dealing with optical images for large territories (more than 10,000 sq. km) is presence clouds and shadows that result having missing values data sets. In this paper, a new approach to classification multi-temporal satellite imagery due proposed. First, self-organizing Kohonen maps (SOMs) are used restore pixel time series imagery. SOMs trained each spectral band separately using nonmissing values. Missing restored through special procedure substitutes input...
Abstract. Winter wheat crop yield forecasting at national, regional and local scales is an extremely important task. This paper aims assessing the efficiency (in terms of prediction error minimization) satellite biophysical model based predictors assimilation into winter models different (region, county field) for one regions in central part Ukraine. Vegetation index NDVI, as well parameters (LAI fAPAR) derived from data WOFOST growth are considered model. Due to very short time series...
For many applied problems in agricultural monitoring and food security it is important to provide reliable crop classification maps national or global scale. Large amount of satellite data for large scale mapping generate a "Big Data" problem. The main idea this paper was comparison pixel-based approaches Ukraine exploring efficiency the Google Earth Engine (GEE) cloud platform solving problem providing high resolution map territory. study carried out Joint Experiment Crop Assessment...
In the paper we propose methodology for solving large scale classification and area estimation problems in remote sensing domain on basis of deep learning paradigm. It is based a hierarchical model that includes self-organizing maps (SOM) data preprocessing segmentation (clustering), ensemble multi-layer perceptrons (MLP) heterogeneous fusion geospatial analysis post-processing. The proposed applied generation high resolution land cover use territory Ukraine from 1990 to 2010 2015.
The water cloud model (WCM) can be inverted to estimate leaf area index (LAI) using the intensity of backscatter from synthetic aperture radar (SAR) sensors. Published studies have demonstrated that WCM accurately LAI if is effectively calibrated. However, calibration this requires access field measures as well soil moisture. In contrast, machine learning (ML) algorithms trained satellite data, even moisture are not available. study, a support vector (SVM) was for corn, soybeans, rice, and...
Fire is one of the most common disturbances in natural ecosystems. The analysis various sources information (official and unofficial) about fires Ukraine (2019–2020) showed a lack timely reliable information. Satellite observation crucial importance to provide accurate, reliable, This paper aims modify index fire danger forest’s FWI by increasing its precision, based on use higher spatial resolution satellite data. A modification method involves utilization soil moisture deficit, addition...
Data limitations and attributability issues due to the full-scale Russian invasion of Ukraine in February 2022 presents continuing challenges assessing production major commodity crops Ukraine. Up-to-date satellite imagery provides evidence rapid changes cropland within temporary occupied territories (TOT) by Russia is world's top producer exporter sunflower and, therefore, monitoring, quantifying areas extremely important. We used Sentinel-1 (S1) synthetic aperture radar (SAR) images...
Modern imaging systems produce a great volume of image data. In many practical situations, it is necessary to compress them for faster transferring or more efficient storage. Then, compression has be applied. If images are noisy, lossless almost useless, and lossy characterized by specific noise filtering effect that depends on the image, noise, coder properties. Here, we considered modern HEIF applied grayscale (component) different complexity corrupted additive white Gaussian noise. It...
Floods are among the most devastating natural hazards in world, affecting more people and causing property damage than any other phenomena. One of important problems associated with flood monitoring is a extent extraction from satellite imagery, since it impractical to acquire area through field observations. This paper presents new method synthetic-aperture radar (SAR) images that based on intelligent computations. In particular, we apply artificial neural networks, self-organizing...
In this paper, we focus on the application of satellite synthetic-aperture radar (SAR) images for discriminating summer crops in Ukraine within JECAM project. Both optical (EO-1/ALI) and SAR (RADARSAT-2) are used order to assess impact adding classification purposes. Three different classifiers, particular neural networks, support vector machine decision trees, applied with networks giving best overall accuracy. It is found that major using sunflower sugar beet classes while there was no...