KLASIFIKASI PENUTUP LAHAN BERBASIS OBJEK PADA DATA FOTO UAV UNTUK MENDUKUNG PENYEDIAAN INFORMASI PENGINDERAAN JAUH SKALA RINCI

Article Sidebar

PDF
Published: Dec 1, 2014
DOI: https://doi.org/10.30536/inderaja.v11i2.3304
Keywords:
UAV aerial photo, Object based classification, Texture information, land cover

Main Article Content

Nurwita Mustika Sari
Dony Kushardono
Indonesian National Institute of Aeronautics and Space (LAPAN)

Abstract

The need of spatial information from detailed-scale remote sensing is increasing. Unmanned Aerial Vehicle or UAV become one of vehicles that is expected to obtain such information. Production of land cover spatial information using UAV photo data requires appropriate method for classification. This study proposes an object-based classification method for land cover based on Haralick texture information namely homogeneity, contrast, dissimilarity, entropy, angular second moment, mean, standard deviation, and correlation. As a comparison method, a conventional land cover-object-based classification is implemented using the same information features, there are brightness, compactness, and density. The result shows that method using texture feature in object-based classification has reached 95.22% accuracy or 17.5% difference that is much better than conventional method that reaches 77.71%.

Article Details

How to Cite
[1]
N. M. Sari and D. Kushardono, “KLASIFIKASI PENUTUP LAHAN BERBASIS OBJEK PADA DATA FOTO UAV UNTUK MENDUKUNG PENYEDIAAN INFORMASI PENGINDERAAN JAUH SKALA RINCI”, INDERAJA, vol. 11, no. 2, pp. 114–127, Dec. 2014.
Issue
Vol. 11 No. 2 (2014)
Section
Articles
Creative Commons License

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.

Author Biography

Nurwita Mustika Sari

The need of spatial information from detailed-scale remote sensing is increasing. Unmanned Aerial Vehicle or UAV become one of vehicles that is expected to obtain such information. Production of land cover spatial information using UAV photo data requires appropriate method for classification. This study proposes an object-based classification method for land cover based on Haralick texture information namely homogeneity, contrast, dissimilarity, entropy, angular second moment, mean, standard deviation, and correlation. As a comparison method, a conventional land cover-object-based classification is implemented using the same information features, there are brightness, compactness, and density. The result shows that method using texture feature in object-based classification has reached 95.22% accuracy or 17.5% difference that is much better than conventional method that reaches 77.71%.