IMPLEMENTASI ALGORITMA CONVOLUTIONAL NEURAL NETWORK UNTUK MENGHITUNG KEPADATAN AYAM MENGGUNAKAN YOU ONLY LOOK ONCE

Abstract

This research aims to apply the Convolutional Neural Network (CNN) algorithm to count the number of chickens in an image using the You Only Look Once (YOLO) method. YOLO is a reliable and real-time object detection algorithm.

This research uses a dataset of chicken images collected from farms. The images are then preprocessed to improve their quality before being used as input for the CNN model. The CNN model is trained using the backpropagation algorithm to minimize the loss function.

The research results showed that the use of YOLO was successful in producing an application for calculating chicken density. Blackbox testing shows that the application created functions well and meets user needs and the results of the matrix evaluation test show an average accuracy of 84%.