Land cover classification using sentinel-1 SAR data

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F61989100%3A27350%2F19%3A10251611" target="_blank" >RIV/61989100:27350/19:10251611 - isvavai.cz</a>

Výsledek na webu

<a href="https://ieeexplore.ieee.org/document/8870125" target="_blank" >https://ieeexplore.ieee.org/document/8870125</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1109/MILTECHS.2019.8870125" target="_blank" >10.1109/MILTECHS.2019.8870125</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Land cover classification using sentinel-1 SAR data

Popis výsledku v původním jazyce

With the development of remote sensing techniques, optical images become more efficient compare to field survey. However, the quality of optical images would influenced by cloud. Radar is known to be very sensitive to vegetation physiognomy and biomass. The sensitivity of synthetic aperture radar (SAR) to the structural features of terrain leads to landcover classification into simple and easily interpreted structural classes. In this paper, the potential of using free of charge Sentinel-1 SAR imagery for land cover mapping in the Moravian-Silesian region, is investigated. The images recorded in 2018 were used for a per-pixel and object-based classification of agricultural land. The per-pixel classification was performed by the maximum likelihood algorithm, the object-based classification then using the support vector machine algorithm. The legend was taken from the Land Parcel Identification System (LPIS) and contained the following three classes - grassland, arable land and a class that involves hop fields, vineyards, and orchards. Post processing of the classification results has been done using the confusion matrix (also known as error matrix) and corresponding overall accuracy and Kappa coefficients of all the classification methods have been calculated. Significantly better results were achieved through object-oriented classification. In both areas of interest, the highest processing and user precision was achieved for the arable land class. (C) 2019 IEEE.

Název v anglickém jazyce

Land cover classification using sentinel-1 SAR data

Popis výsledku anglicky

With the development of remote sensing techniques, optical images become more efficient compare to field survey. However, the quality of optical images would influenced by cloud. Radar is known to be very sensitive to vegetation physiognomy and biomass. The sensitivity of synthetic aperture radar (SAR) to the structural features of terrain leads to landcover classification into simple and easily interpreted structural classes. In this paper, the potential of using free of charge Sentinel-1 SAR imagery for land cover mapping in the Moravian-Silesian region, is investigated. The images recorded in 2018 were used for a per-pixel and object-based classification of agricultural land. The per-pixel classification was performed by the maximum likelihood algorithm, the object-based classification then using the support vector machine algorithm. The legend was taken from the Land Parcel Identification System (LPIS) and contained the following three classes - grassland, arable land and a class that involves hop fields, vineyards, and orchards. Post processing of the classification results has been done using the confusion matrix (also known as error matrix) and corresponding overall accuracy and Kappa coefficients of all the classification methods have been calculated. Significantly better results were achieved through object-oriented classification. In both areas of interest, the highest processing and user precision was achieved for the arable land class. (C) 2019 IEEE.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10500 - Earth and related environmental sciences

Projekt

—

Návaznosti

S - Specificky vyzkum na vysokych skolach

Rok uplatnění

2019

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

ICMT 2019 - 7th International Conference on Military Technologies, Proceedings

ISBN

978-1-72814-593-8

ISSN

—

e-ISSN

—

Počet stran výsledku

5

Strana od-do

—

Název nakladatele

IEEE

Místo vydání

Piscataway

Místo konání akce

Brno

Datum konání akce

30. 5. 2019

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—