Title: The evaluation of the usage of the fuzzy algorithms in increasing the accuracy of the extracted land use maps
Authors: Mojtaba Pirnazar; Hafez Hasheminasab; Arash Zand Karimi; Kaveh Ostad-Ali-Askari; Zahra Ghasemi; Majedeh Haeri-Hamedani; Elham Mohri-Esfahani; Saeid Eslamian
Addresses: Department of Remote Sensing, Tabriz University, Tabriz, Iran ' Department of Remote Sensing, Tabriz University, Tabriz, Iran ' Department of Remote Sensing, Tabriz University, Tabriz, Iran ' Department of Civil Engineering, Isfahan (Khorasgan) Branch, Islamic Azad University, Isfahan, Iran; Islamic Azad University, Khorasgan Branch University Blvd., Arqavanieh, Jey Street, Isfahan, Iran P.O. Box: 81595-158 ' Department of Remote Sensing, Islamic Azad Yazd University, Iran ' Water Engineering, Isfahan University of Technology, Isfahan, Iran ' Engineering in Irrigation and Drainage, Isfahan University of Technology, Isfahan, Iran ' Water Engineering Department, Isfahan University of Technology, Isfahan, Iran
Abstract: Within this paper, we evaluate the accuracy of three methods of classification including: object-oriented algorithms of the satellite images classification without the use of fuzzy algorithms, algorithm based on fuzzy algorithms, and pixel-based algorithms The accuracy of each method obtained by comparing the results with pixel-based algorithm in land use/land cover classification in Maragheh County. To reach this goal, AVNIR2 sensor images that generated from ALOS satellite were used to classify land use. The results obtained from the methods indicated that the classifications which produced by object-oriented classification method were more accurate than that of pixel-based method. The accuracy of fuzzy knowledge-based method was 93.28%. However the accuracy of the object-oriented method without using of the fuzzy algorithms and the pixel-based algorithm method were 88.06% and 83.79% respectively. According to these results, using higher spatial resolution images along with proper algorithms for extracting of features of land use classes is recommended to environmental researches.
Keywords: remote sensing; object-oriented classification methods; pixel base; fuzzy algorithms; land use map.
International Journal of Global Environmental Issues, 2018 Vol.17 No.4, pp.307 - 321
Received: 27 Sep 2016
Accepted: 30 Jan 2017
Published online: 21 Jun 2018 *