Jurnal Ilmiah ESAI
Vol 8 No 3 (2014)

Detecting Resemblance Of Orchid Plant Image Through Support Vector Machine (SVM) Of Kernel Linear Method

Dewi Kania Widyawati (Unknown)
Zuriati Zuriati (Unknown)



Article Info

Publish Date
29 Jun 2018

Abstract

The research dealt with detecting resemblance of orchid plant image through Support vector machine (SVM) of Kernel Linear method. With one versus rest modeling, the images were taken by using single type camera Canon S550D. The use of trial data and test data varied into for ratio types namely : 50% trial data - 50% test data , 60% trial data - 40% test data , 70% trial data - 30% test data , and 80% trial data - 20% test data. The extract of texture features was done with combining operator circular neighborhood (8,1) and (8,2) and concatenation done through fuzzyfication. The research aimed to (1) design a program to detect the resemblance of orchid plant image (2) implement Support Vector Machine kernel Linear method with one versus Rest model to identify the image of orchid plants both with and without flowers (3) analyze distribution level of accuracy of the four trial and test data examined from each specimen. (4) Analyze resemblance of orchid plant image through Support Vector Machine kernel Linear with one versus Rest model. The research was carried out through: (1) collecting the image and praposes (2) extracting the textures, (3) classifying the Support Vector Machine kernel Linear, (4) data testing and (5) evaluating classification result. The main target of the research is to find out a system to detect the resemblance of orchid plants both with and without flower.Keywords: circular neighborhood, one versus rest, Support Vector Machine kernel Linear

Copyrights © 2014






Journal Info

Abbrev

ESAI

Publisher

Subject

Economics, Econometrics & Finance

Description

Jurnal ESAI menerbitkan artikel penelitian dan studi konseptual Ekonomi. Jurnal ini diterbitkan oleh Jurusan Ekonomi dan Bisnis, Politeknik Negeri Lampung. Artikel penelitian yang dikirimkan ke jurnal Esai ini akan direview oleh 2 (dua) reviewer. Artikel penelitian yang diterima akan tersedia online ...