Penerapan Metode Deep Learning dengan Model CNN untuk Mendukung Sistem Deteksi Penghitungan Benur Udang Vaname

Anggia Putri, Tri Farin Meydiantika (2025) Penerapan Metode Deep Learning dengan Model CNN untuk Mendukung Sistem Deteksi Penghitungan Benur Udang Vaname. Diploma thesis, Politeknik Negeri Jember.

	Text (Abstrak) Abstrak.pdf - Submitted Version Available under License Creative Commons Attribution Share Alike. Download (234kB)
	Text (Bab 1 Pendahuluan) Bab 1. Pendahuluan.pdf - Submitted Version Available under License Creative Commons Attribution Share Alike. Download (237kB)
	Text (Daftar Pustaka) Daftar Pustaka.pdf - Submitted Version Available under License Creative Commons Attribution Share Alike. Download (192kB)
	Text (Laporan Lengkap) Laporan Lengkap.pdf Restricted to Registered users only Download (5MB) | Request a copy

Abstract

Vaname shrimp farming is a strategic sector in Indonesian fisheries, but the manual counting process is often an obstacle in production efficiency and accuracy. This research aims to develop an automatic vaname shrimp fry detection and counting system using deep learning method with Convolutional Neural Network (CNN) model and YOLOv8 architecture. The dataset used was 757 augmented images from 300 original images, with the training process carried out on the Roboflow platform for 260 epochs. The training results showed good model performance with a mean average precision (mAP) value of 92%, precision of 89.8%, recall of 90.5%, and average loss of 1.3. Model testing resulted in an accuracy of 85.2% based on the confusion matrix. The model was able to efficiently detect and count vaname shrimp fry in less than five minutes, making it a potential solution in the automation of fry monitoring systems to support digital transformation in the aquaculture sector.

Actions (login required)

View Item