Perbandingan Object Oriented Classification dan Maximum Likelihood Classification pada Pemetaan Penutupan Lahan di Kabupaten Gayo Lues

Muhammad Rusdi

Abstract


Comparison of object Oriented Classification With Maximum Likelihood Classification on Land Cover Maaping in District of Gayo Lues

ABSTRACT. Image classification method used critically effects the classified image produced as well as its accuracy. One of the recently introduced classification methods is Object-Oriented Classification (OOC). The fundamental difference between this method and that of Maximum Likelihood Classification (MLC) lies on the basic unit of image analytical prosses, i.e., image object or segment, and not on a single pixed. The new classification method uses a segmentation procedure with a hierarchical approach which allows an addition of an object  characteristic to other information related  to the classified object, such as from, texture, and context. This object or segment is formed because the smallest region has a larger area than the image pixel. In this study, MLC and OOC classification methods were compered using Landsat ETM+ image for Gayo Lues District of Aceh. The objective of this study was to map out land cover/ landuse classes of the research area and compare the result and its accuracy using the MLC and OOC classification methods. Result  of the study showed that a higher, more detailed, hierarchical land cover/ landuse class with a higher accuracy was obtained using OOC classification method which corresponds better to the contextual condition in the field compared to that of the pixel- based classification method.

Keywords


object oriented classification; maximum likehood classification; segmentation; hierarchical

Full Text:

PDF

Refbacks

  • There are currently no refbacks.



Creative Commons LicenseISSN: 1411-4623E-ISSN: 2460-4534
Copyright© 1997-2018 | ISSN: 1410-3389 | EISSN: 2597-9973 
Jurnal Agrista is licensed under a Creative Commons Attribution 4.0 International License.

 

Published by: 
The Faculty of AgricultureSyiah Kuala University 
Jl. Tgk. Hasan Krueng Kalee No. 3, Kopelma Darussalam, 
Banda Aceh, 23111, Indonesia. 
Phone: +62-81383736633 
Email: agrista@unsyiah.ac.id


Online Submissions & Guidelines Editorial Policies | Contact Statistics Indexing | Citations