Search academic texts

Information × Registration Number 0221U106751, 0121U111242 , R & D reports Title Geospatial models and information technologies of satellite monitoring of smart city problems popup.stage_title Head Shelestov Andrii Yu., д.т.н. Registration Date 19-12-2021 Organization National Technical University of Ukraine "Igor Sikorsky Kyiv Polytechnic Institute" popup.description2 Report SRW: 62 р., there are: 2 chapters, 12 figures, 3 tables, 94 sources. Research object — geospatial data classification methods The subject of research - intelligent models and information technologies of classification on the basis of high spatial resolution satellite data Purpose of work - to develop and implement an algorithm for classifying the land cover on the basis of high spatial resolution satellite data, improvement of convolutional neural networks with subsequent filtering of the obtained map and extraction from it of building density by city quarters. Research methods - neural network models of land cover classification based on satellite data, software tools for working with geospatial data Results: Information features based on satellite data of high spatial resolution for different land cover types for land cover classification of the city of Kyiv are singled out and investigated. The architectures of neural networks used in the world for similar classification tasks are studied, the architecture of the neural network for segmentation and classification of satellite data of high spatial resolution for the city of Kyiv is developed. Using additional probabilistic information on the recognition of land cover classes, a method of filtering the obtained land cover classification map was developed. A method for estimating the buildings density within urban neighborhoods based on the obtained map of land cover classification has been developed. Metrics for the accuracy estimation of the obtained results are determined. All these methods and metrics are implemented in the Python programming language. Within the framework of the project, an article on “Air Quality Estimation in Ukraine Using SDG 11.6.2 Indicator Assessment” was published in the international journal Remote Sensing, which is indexed in SCOPUS and belongs to quartile Q1. Keywords: CONVOLUTIONAL NEURAL NETWORK, PROBABILICTIC CLASSIFICATION, LAND COVER MAP, URBAN ATLAS, SMART CITY Product Description popup.authors Drozd Sofiia Yu. Karpenko Oleksandr S Kolotii Andrii V. Lavreniuk Alla M Mikava Polina V. Yailymova Hanna O. popup.nrat_date 2021-12-19 Close