Fruit Classification Comparison Based on CNN and YOLO

Abstract

Today, agriculture has become more of science than art. The demand for fresh produce is growing, while availability of field workers is diminishing. Also, Farmers keep investing in land, irrigation, and labor, and much of the fruit is left hanging on the trees because there are not enough people to pick it when it is ripe. Automation in the agricultural field using modern information and Communication Technology can increase the efficiency, reliability, and precision; and reduces the need of human intervention. Thus, to meet solution to this problem is by reducing the costs for farmers and increasing the yield of crops which can be achieved by classifying of fruit by using modern technique in the field of Deep Neural Networking that involves CNN, YOLO etc. which provides a helping hand to farmers to optimize their farming patterns and boost yields by allowing them to grow taller fruit trees that can be harvested by classifying based on the condition of fruit and saving significant labour cost. In conclusion, the proposed methods are highly acceptable and recognizing the necessity, and unresolved classifications.

Penerapan Algoritma CNN Menggunakan Framework Yolo Untuk Deteksi Objek Produk Di Perusahaan Manufaktur

“…Salah satu pengolahan citra digital adalah algoritma Convolutional Neural Network (CNN) (Li, 2022). Algoritma CNN adalah suatu jaringan syaraf tiruan dengan menggunakan metode feed-forward (Raj, 2021). Arsitektur CNN meliputi input citra, convolutional layer, pooling layer atau max pooling, activation function dan fully connected layer (Bhatt, 2021) (Yedla, 2020).…”

“…fully connected layer untuk menghubungkan semua fungsi aktivasi dan lapisan konvolusi. CNN dapat mengenali dan mengklasifikasikan gambar melalui pola citra yang terdeteksi dengan akurasi yang baik (Narvekar, 2020) (Raj, 2021). Secara keseluruhan CNN tidak berbeda jauh dengan neural network biasa.…”