Open AccessThe Implementation of Deep Learning Using Convolutional Neural Network to Classify Based on Stomata Microscopic Image of Curcuma Herbal Plants
Author(s) -
U Andayani,
Imam Bagus Sumantri,
Andes Pahala,
M A Muchtar
Publication year - 2020
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/851/1/012035
Subject(s) - curcuma , convolutional neural network , artificial intelligence , artificial neural network , computer science , pattern recognition (psychology) , genus , deep learning , traditional medicine , botany , biology , medicine
There are so many types of herbal plants that come from the same genus. The similarity of features in herbal plants makes it difficult to distinguish. Especially in the pharmaceutical field, which is very risky for making mistakes. The part of plants that is often used as ingredients for herbal medicines is the leaves, so research on leaves and their constituent organelles is very important for the pharmaceutical world. Therefore, an approach is needed to identify the types of organelles present in the leaves, where the organelles most frequently studied are the stomata. So, a neural network approach is needed to distinguish plant characteristics in the same genus. In this research there are 2 species of plants in the Curcuma genus, namely turmeric and ginger. This research was conducted by implementing Deep Learning using Convolutional Neural Network (CNN) with the help of the Gabor filter process and feature extraction using the Gray Level Co-occurrence Matrix (GLCM). The study was conducted using 160 microscopic images of Curcuma herbal plants as training data with training accuracy of 93.1% and test data of 40 images with an accuracy of 92.5%.