Klasifikasi Varietas Buah Kiwi dengan Metode Convolutional Neural Networks Menggunakan Keras

Abstract

Kiwi fruit is known as a fruit rich in benefits because it contains many nutrients, sources as well as high antioxidants. In Indonesia, there are two varieties of kiwi fruit sold in the market, namely green kiwi and golden kiwi and there is one more variety, namely red kiwi. The content of the three varieties is different and the price is also different. Gold kiwi has the highest nutritional content so that the price is above other kiwi varieties, but from the outside the appearance of this kiwi fruit at a glance is the same and many people do not recognize the kiwi variety they will buy even though these three kiwi varieties have different tastes and nutritional content. For this reason, the researcher proposes a classification system for kiwi fruit varieties using the hard CNN method. The CNN method is one of the deep learning methods that can be used to recognize and classify an object in a digital image. Then the preprocessing process is carried out using labeling on the data. Then the CNN architecture is designed with Input containing 320x258x3 neurons. The data was then trained using 25 epochs with an accuracy rate of 0.98. Then the test data using test data get an average accuracy value of 0.987, while for precision and recall it is also the same at 0.987