Klasifikasi Rasa Berdasarkan Citra Buah Menggunakan Algoritma Convolutional Neural Network dengan Teknik Identitas Ganda

Refiani, Pintanarum (2021) Klasifikasi Rasa Berdasarkan Citra Buah Menggunakan Algoritma Convolutional Neural Network dengan Teknik Identitas Ganda. Undergraduate Thesis thesis, Institut Teknologi Telkom Purwokerto.

	Text COVER.pdf Download (536kB)
	Text ABSTRACT.pdf Download (6kB)
	Text ABSTRAK.pdf Download (107kB)
	Text BAB I.pdf Download (375kB)
	Text BAB II.pdf Download (677kB)
	Text BAB III.pdf Download (981kB)
	Text BAB IV.pdf Restricted to Registered users only Download (577kB) | Request a copy
	Text BAB V.pdf Download (9kB)
	Text DAFTAR PUSTAKA.pdf Download (134kB)
	Text LAMPIRAN.pdf Restricted to Registered users only Download (382kB) | Request a copy

Abstract

Multiple identities whose are a natural thing to find in the real world, for example, transgender is one of them. In this study, distinguishing fruits that have more than one type of taste, such as lemon, has a dual identity, namely sour and salty. For deep learning categorization, often the training dataset population has different objects from one another. In the taste contained in the fruit, we can exploit it to see the behavior of the Convolutional Neural Network (CNN) in recognizing objects that have more than one identity. The training phase in this study is that in two different taste categories, several images of the same fruit are intentionally inserted into these two categories. The accuracy of this categorization is then compared with the accuracy of the pure categorization, there are no twin or multiple objects in each category. From the simulation results, the accuracy of the model when there are multiple identities is 95%, while the pure one is 98%. From this research, the best total accuracy in identifying the taste of 27 types of fruit is the 7B model with a total accuracy of 98%. Keywords: Augmentation, Convolutional Neural Network, Deep Learning, Fruit, Python, Taste.

Actions (login required)

View Item