Monitoring the spread of Sosnowskyi’s hogweed using a random forest machine learning algorithm in Google Earth Engine

Yifter T.T., Razoumny Y.N., Orlovsky A.V., Lobanov V.K.
 pdf (2791K)

Examining the spectral response of plants from data collected using remote sensing has a lot of potential for solving real-world problems in different fields of research. In this study, we have used the spectral property to identify the invasive plant Heracleum sosnowskyi Manden from satellite imagery. H. sosnowskyi is an invasive plant that causes many harms to humans, animals and the ecosystem at large. We have used data collected from the years 2018 to 2020 containing sample geolocation data from the Moscow Region where this plant exists and we have used Sentinel-2 imagery for the spectral analysis towards the aim of detecting it from the satellite imagery. We deployed a Random Forest (RF) machine learning model within the framework of Google Earth Engine (GEE). The algorithm learns from the collected data, which is made up of 12 bands of Sentinel-2, and also includes the digital elevation together with some spectral indices, which are used as features in the algorithm. The approach used is to learn the biophysical parameters of H. sosnowskyi from its reflectances by fitting the RF model directly from the data. Our results demonstrate how the combination of remote sensing and machine learning can assist in locating H. sosnowskyi, which aids in controlling its invasive expansion. Our approach provides a high detection accuracy of the plant, which is 96.93%.

Keywords: Heracleum Sosnowski Manden, invasive plants, Google Earth Engine, machine learning, random forest
Citation in English: Yifter T.T., Razoumny Y.N., Orlovsky A.V., Lobanov V.K. Monitoring the spread of Sosnowskyi’s hogweed using a random forest machine learning algorithm in Google Earth Engine // Computer Research and Modeling, 2022, vol. 14, no. 6, pp. 1357-1370
Citation in English: Yifter T.T., Razoumny Y.N., Orlovsky A.V., Lobanov V.K. Monitoring the spread of Sosnowskyi’s hogweed using a random forest machine learning algorithm in Google Earth Engine // Computer Research and Modeling, 2022, vol. 14, no. 6, pp. 1357-1370
DOI: 10.20537/2076-7633-2022-14-6-1357-1370
URL: http://crm-en.ics.org.ru/journal/article/3281/
Creative Commons License This work is licensed under a Creative Commons Attribution-NoDerivs 3.0 Unported License.

Indexed in Scopus

Full-text version of the journal is also available on the web site of the scientific electronic library eLIBRARY.RU

The journal is included in the Russian Science Citation Index

The journal is included in the RSCI

International Interdisciplinary Conference "Mathematics. Computing. Education"

Copyright © 2009–2024 Institute of Computer Science