Institusion
Institut Teknologi Perusahaan Listrik Negara
Author
Meidilaga, Gema Naufal
Affandi, Riki Ruli
Manjawakang, Abdul Haris
Subject
Teknik Informatika
Datestamp
2023-06-21 08:27:16
Abstract :
In the growth of rice plants there are abnormalities in growth and delays in
anticipating which will lead to sub-optimal yields. This is caused by blas disease and blight.
To assist in the fast handling, this study designed a website-based expert system with the
Faster Region Convolutional Neural Network method approach which aims to help solve
problems because this method functions to process image data and the image data used is
leaf image data on rice plants. This study resulted in an accuracy of 99.32% in identifying
and classifying diseases using leaf imagery. Evaluation using confusion matrix results in
accuracy of 94%, precision of 100%, recall of 74%, and f1-score of 85%.