Search academic texts

Information × Registration Number 0226U002346, (0124U001745) , R & D reports Title Improving the methodology of spatial identification of crops using remote sensing popup.stage_title Удосконалення методів збирання та обробки супутникових знімків для визначення посівів сільськогосподарських культур та видів земельних угідь Head Melnyk Denis M., Кандидат економічних наук Registration Date 23-02-2026 Organization Institut of Land Use NAAS Ukraine popup.description1 The aim of the project is to develop ways to improve approaches to the spatial identification of crops using remote sensing due to the need to improve the quality and efficiency of collecting and processing data on agricultural land use under martial law and post-war reconstruction of the country. Due to the deterioration of conditions for statistical accounting, remote sensing using artificial intelligence and geographic information technologies is the most important tool for collecting and analysing data on actual land use. popup.description2 The study analyzed the difficulties associated with the identification of agricultural crops on satellite images. One of the main obstacles is atmospheric conditions that distort the spectral characteristics of objects. Additional difficulties arise due to the seasonal dynamics of crop growth, which changes their visual characteristics throughout the year. To improve the identification results, it is necessary to adapt satellite data taking into account the phases of crop development, select relevant spectral channels and indices, and select the appropriate spatial resolution, which will allow achieving high classification accuracy. A key role in crop monitoring is played by vegetation indices – in particular, NDVI and EVI, which, based on reflection in the red and near-infrared spectra, allow estimating plant biomass and identifying periods of active growth or stress. The choice of index depends on the type of crop, its structure and stage of vegetation. This approach allows for regular monitoring of crop conditions, predicting yields and detecting changes associated with climatic or anthropogenic factors. To improve the efficiency of observations in areas where hostilities are underway, an automated system for classifying agricultural crop areas using the Google Earth Engine platform has been implemented. A comparative analysis of data for 2021–2024 has been conducted, which allows us to assess the dynamics of changes. Due to the limitations of field research, satellite technologies and machine learning algorithms are an alternative and reliable way to monitor agricultural activities in wartime conditions. Product Description popup.authors Ibatullin Shamil I. Dorosh Yosyp M. Sakal Oksana V. Kharytonenko Roman A. Trokhymchuk Andrii A. Saliuta Victoriia A. Solomakha Iryna V. Olishevskyi Valerii V. popup.nrat_date 2026-02-23 Close