Institusion
Institut Teknologi Perusahaan Listrik Negara
Author
TAMPUBOLON, FACHRI YUSZA PARLINGGOMAN
Rusjdi, Darma
Djunaidi, Karina
Subject
Teknik Elektro
Datestamp
2023-06-07 07:00:52
Abstract :
Chocolate is a food that is in demand from people of all ages, besides being delicious,
chocolate also has a variety of nutritional content that has various benefits such as
improving brain function, chocolate itself is the grinding process of cocoa beans on the
cocoa plant. The problem faced is that cocoa plants on the surface of the cocoa pods have
various types of diseases such as black pod, frosty pod and pod borer, these diseases can
reduce the quality of the cocoa beans produced and can also create a pandemic around the
cocoa plantation which results in crop failure. good technology is needed to overcome this
problem such as calcification of the type of disease on the skin of cocoa pods using the
convolutional neural network method, having good quality and a fairly good level of
accuracy. In the process, modeling will be made on objects carried out on hard tensorflow
software which in the future results The results obtained from the model process will be
included in the application for calcification of cocoa pods. This study uses four types of
classification, namely healthy class, black pod frosty pod and pod borer. The results
obtained in this study were 0.9569 on the tensorflow object train and 0.86 on the valitadion.
the prediction of the tensorflow model obtained is 86.75%. from the results obtained,
developments can be made for the manufacture of detection tools for cocoa pod skin
disease