Mesterséges intelligencia alapú osztályozó eljárások alkalmazása nagytájszintű területhasználat változás térképezésére

Gudmann András: Mesterséges intelligencia alapú osztályozó eljárások alkalmazása nagytájszintű területhasználat változás térképezésére. In: Móra Akadémia : szakkollégiumi tanulmánykötet 10., (10). pp. 239-253. (2022)

[thumbnail of moraakademia_010_239-253.pdf] Cikk, tanulmány, mű
moraakademia_010_239-253.pdf

Download (2MB)

Abstract

The land use change properly shows the transformation of a certain area. This study aims to create and analyse land-use maps from Hungary in 4 different acquisition times (from 1986 to 2017). For the land-use map creation, Landsat mosaics with texture data were created. Two different classification algorithm was tested: artifical neural network and random forest. The test shows that the random forest procedure was more suitable for this task, beacuse its speed and high accuracy. After the procedure selection random forest was applied for the classification task. For reference data CORINE Land Cover databases (CLC90, CLC00, CLC06, CLC12, CLC18) were applied. The accuracy assesment had shown the land-use maps high overall accuracy (78,69 – 90,67%). A land-use map were compared with each other, and the main change flows were determined. According the land-use change results, the main land-use processes were the intensification and extensification of agriculture, the urbanisation and afforestation. The intensification and extensification of agriculture were high in the whole country area during the investigated period. The urbanisation was significant in the main agricultural areas and the afforestation in mountains. Thanks of these processes the area of the artificial surfaces were increased with 217 thousand hectares and the size of the forests with 94 thousand hectare. However, the size of the non-irrigated arable lands and with that, the agricultural lands were decreased with significant (307 thousand and 384 thousand hectare).

Item Type: Book Section
Other title: Application of artificial intelligence-based classification algorithms to map large-scale land use change
Journal or Publication Title: Móra Akadémia : szakkollégiumi tanulmánykötet 10.
Date: 2022
Volume: 10
ISSN: 2064-809X
Page Range: pp. 239-253
Series Name: Móra Akadémia
Language: Hungarian, English
Publisher: Szegedi Tudományegyetem Móra Ferenc Szakkollégiuma
Place of Publication: Szeged
Related URLs: http://acta.bibl.u-szeged.hu/76360/
Uncontrolled Keywords: Területhasználat-változás
Additional Information: Bibliogr.: p. 250-252. és a lábjegyzetekben ; ill. ; összefoglalás angol nyelven
Subjects: 01. Natural sciences
01. Natural sciences > 01.05. Earth and related environmental sciences
Date Deposited: 2022. Aug. 08. 11:41
Last Modified: 2022. Aug. 08. 11:41
URI: http://acta.bibl.u-szeged.hu/id/eprint/76378

Actions (login required)

View Item View Item