Implementasi Sistem Pengenal Berbasis Citra dan IoT Untuk Deteksi Produktivitas Ayam Petelur

Anggoro, Cevin Pradipta (2025) Implementasi Sistem Pengenal Berbasis Citra dan IoT Untuk Deteksi Produktivitas Ayam Petelur. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5009211133_Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (10MB) | Request a copy

Abstract

Peningkatan produktivitas pada peternakan ayam petelur merupakan faktor untuk mendukung ketahanan pangan nasional. Namun, metode pemantauan produktivitas yang masih bersifat manual memiliki banyak keterbatasan, seperti ketergantungan pada tenaga kerja manusia, serta risiko kesalahan pencatatan yang berdampak pada manajemen pakan yang kurang baik. Penelitian ini bertujuan untuk mengembangkan sistem pengenal berbasis citra dan Internet of Things (IoT) untuk deteksi produktivitas ayam petelur secara otomatis dan real time. Sistem menggunakan kamera OV2640 pada modul ESP32-CAM untuk menangkap gambar telur dan label kandang, kemudian memanfaatkan algoritma YOLO untuk mendeteksi objek telur serta membaca label kandang. Data hasil deteksi diolah menjadi file excel dan dikirim ke platform Blynk melalui ESP32 agar dapat dipantau secara daring. Hasil pengujian menunjukkan bahwa model YOLO untuk deteksi telur pada pengujian 1 memiliki performa yang sangat baik dengan akurasi 98%, presisi 98%, sensitivitas 100%, dan skor F1 99%. Model label kandang pada pengujian 1 juga menunjukkan kinerja yang cukup baik dengan akurasi 98%, presisi 89%, sensitivitas 100%, spesifisitas 97%, dan skor F1 94%. Pada pengujian 2, model deteksi telur tetap menunjukkan performa tinggi dengan akurasi 97%, presisi 97%, sensitivitas 100%, dan skor F1 99%. Model label kandang pada pengujian 2 juga menunjukkan kinerja yang cukup baik dengan akurasi 98%, presisi 79%, sensitivitas 95%, spesifisitas 98%, dan skor F1 85%. Sistem IoT yang dikembangkan menunjukkan performa yang andal dengan waktu delay berkisar antara 1,86 hingga 6,10 ms, throughput berkisar antara 72,80 hingga 238,31 kbps, jitter berkisar antara 2,13 hingga 6,97 ms, serta tidak ditemukan adanya indikasi packet loss dengan packet loss 0%. Dengan demikian, implementasi sistem ini dapat digunakan untuk mendeteksi produktivitas ayam petelur secara real time, dan meningkatkan manajemen pakan ternak.
===================================================================================================================================
Improving productivity in egg-laying chicken farms is a key factor in supporting national food security. However, manual productivity monitoring methods have many limitations, such as dependence on human labor and the risk of recording errors that can lead to poor feed management. This study aims to develop an image-based recognition system and Internet of Things (IoT) for the automatic and real time detection of egg-laying chicken productivity. The system uses an OV2640 camera on the ESP32-CAM module to capture images of eggs and cage labels, then utilizes the YOLO algorithm to detect egg objects and read cage labels. The detection data is processed into an excel file and sent to the Blynk platform via ESP32 so that it can be monitored online. The test results show that the YOLO model for egg detection in test 1 performed very well with an accuracy of 98%, precision of 98%, sensitivity of 100%, and an F1 score of 99%. The cage label model in Test 1 also showed satisfactory performance with an accuracy of 98%, precision of 89%, sensitivity of 100%, specificity of 97%, and an F1 score of 94%. In Test 2, the egg detection model continued to show high performance with an accuracy of 97%, precision of 97%, sensitivity of 100%, and an F1 score of 99%. The cage label model in Test 2 also showed satisfactory performance with an accuracy of 98%, precision of 79%, sensitivity of 95%, specificity of 98%, and an F1 score of 85%. The developed IoT system demonstrated reliable performance with a delay time ranging from 1.86 to 6.10 ms, throughput ranging from 72.80 to 238.31 kbps, jitter ranging from 2.13 to 6.97 ms, and no indication of packet loss with 0% packet loss. As a result, the implementation of this system can be used to detect the productivity of laying hens in real time and improve livestock feed management.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Produktivitas Ayam Petelur, Pemrosesan Citra, Deteksi Citra, IoT, Laying Hen Productivity, Image Processing, Image Detection, IoT
Subjects:	S Agriculture > SF Animal culture T Technology > T Technology (General) > T57.5 Data Processing T Technology > T Technology (General) > T58.8 Productivity. Efficiency T Technology > TA Engineering (General). Civil engineering (General) > TA1637 Image processing--Digital techniques. Image analysis--Data processing.
Divisions:	Faculty of Industrial Technology and Systems Engineering (INDSYS) > Physics Engineering > 30201-(S1) Undergraduate Thesis
Depositing User:	Cevin Pradipta Anggoro
Date Deposited:	04 Aug 2025 12:03
Last Modified:	04 Aug 2025 12:03
URI:	http://repository.its.ac.id/id/eprint/126838

Actions (login required)

View Item