Oil palm age classification on satellite imagery using fractal-based combination

Soffiana Agustin, Putri Aisyiyah Rakhma Devi, Deni Sutaji, Nuniek Fahriani

Research output: Contribution to journalArticlepeer-review

8 Citations (Scopus)

Abstract

Palm oil tree is one of the sources of vegetable oil and it has become the world’s biggest need of vegetable oil comparing to other plant. The use of satellite imagery to manage the plantation is very helpful for the stakeholders in supervising the development. To find out the age of oil palm trees from the satellite imagery is done by using the extraction of fractal based texture features. This study purpose an feature extraction methods to recognize the age of palm oil tree on the imagery of panchromatic icons using algorithm of segmentation-based fractal texture analysis (SFTA) which is combined with local features and local binary pattern (LBP). The classification will be done by using multi-layer perception method into four classes, which are the age of adult, young, old palm, and the ones which are not palm. Extraction feature is done in fractal-based using SFTA algorithm, giving accurate result of 72.5 %. The combination of feature extraction using SFTA with the local features give accuracy of 74.5 % and the combination of SFTA and LBP gives the accuracy of 76 %.

Original languageEnglish
Pages (from-to)18-26
Number of pages9
JournalJournal of Theoretical and Applied Information Technology
Volume89
Issue number1
Publication statusPublished - 15 Jul 2016
Externally publishedYes

Keywords

  • Classification
  • Ikonos-Panchromatic
  • Local Binary Pattern (LBP)
  • Local Features
  • Oil Palm Age
  • Segmentation Based Ffractal Fexture Analysis (SFTA)

Fingerprint

Dive into the research topics of 'Oil palm age classification on satellite imagery using fractal-based combination'. Together they form a unique fingerprint.

Cite this