Klasifikasi Penyakit pada Buah Kakao Menggunakan Metode Faster Region Convolutional Neural Network

Abstract

Cocoa (Theobroma cacao L.) is a perennial plant in the form of a tree that originates from South America. From the seeds of this plant, it is typically processed into a consumable product commonly known as chocolate. Indonesia is the world's 3rd largest cocoa producer and exporter after Côte d'Ivoire and Ghana. One of the parameters of good quality cocoa fruit is the absence of diseases attached to the cocoa fruit. The most problem that often experienced by cocoa farmers is the presence of pests or diseases that attack cocoa plants. In research (Malik, 2021), the types of diseases on cocoa fruit can be fruit rot (Phytophtora palmivora), anthracnose (Colletotrichum gloeosporioides), black spot (Helopeltis sp) and cocoa fruit borer (Conopomorpha cramerella). This research produces a system that can detect disease in cocoa fruit by looking at the colour of the cocoa fruit. The types of diseases that can be classified in this research are fruit rot (Phytophtora palmivora), anthracnose (Colletotrichum gloeosporioides), black spot (Helopeltis sp) and cocoa pod borer (Conopomorpha cramerella). The data used in this study amounted to 1000 data which were then divided into 800 training data, 100 validation data, and 100 test data. After testing, this research resulted in an accuracy of 95%.