Rhizomes Classification Using Deep Neural Network Based on Odor Maimunah, Seta Samsiana
Universitas Muhammadiyah Magelang
Universitas Islam 45 Bekasi
Abstract
Rhizomes are kind of herbal plant that has many benefits for health. There are many kinds of rhizomes in Indonesia, such as temuireng, temulawak and temumangga, which have similarities in color, shape and odor. Therefore, for ordinary people, they are difficult to be identified. In this reasearch, some rhizomes were classified into three classes, namely the class of temuireng, temulawak and temumangga based on their odor. The odor of a rhizome was captured by a TGS2600 sensor which was connected to the raspberry phi 3B in the voltage value. The voltage data ware taken in 20 minutes for each class before classification process using deep neural network. Results showed that deep neural network could classify the type of a rhizome based on its odor with 98.8% accuracy
Keywords: deep neural network, classification, rhizomes
