Aplikasi Mobile Untuk Identifikasi Tumbuhan Obat Menggunakan Local Binary Patterndengan Klasifikasi Probabilistic Neural Network
Abstract
This research proposed a new mobile application for medicinal plants identification. Local Binary Pattern (LBP) is used to extract image features and Probabilistic Neural Network (PNN) is used to classify medicinal plants on mobile phone. LBPP,Rriu2 was used as LBP descriptor in this research, which uses rotation invariant and uniform patterns of image texture. The aim of this research was to implement this application in open source Android 2.3 (Gingerbread) operating system. The data used is the medicinal leaf image from Botanical Garden, Bogor Indonesia which consist of 15 species with 10 variations for each species. The experiment showed that the accuracy of identifications is 31.11% and the computation time is 50.02 seconds.The experimental result show the accuracy is still low, so this research need to be explored further to improve the accuracy. This new mobile application is useful to help user in plant identification.
Collections
- UT - Computer Science [2323]