Detection of Palm Fruit Maturity Using ConvolutionalNeural Network Method

Abstract

Palm oil has an important role as a source of foreign
exchange in the economy in Indonesia. Oil palm is one of the
vegetable oil-producing plants that has the highest economic
value compared to other crops such as soybeans, olives,
coconuts or sunflowers. Palm oil quality is also influenced by
water content, dirt content, free fatty acid content and the
level of maturity of the palm fruit. Maturity of palm fruit is a
very important factor in determining the quality of crude oil
produced by palm fruit. In determining the maturity of oil
palm, sorting is necessary to get quality palm fruit with the
appropriate level of maturity. The use of image processing
technology (ImageProcessing) can facilitate the process of
analyzing objects. Meanwhile, the implementation of deep
learning using the Convolutional Neural Network method
can help identify the maturity level of oil palm fruit with a
high level of accuracy. The results showed a very good
effectiveness with an accuracy reaching 99% and a precision
level reaching 99.8%.