Please use this identifier to cite or link to this item:
http://repository.ipb.ac.id/handle/123456789/120101| Title: | Integrasi MobileNet-SSD dan Isolation Forest untuk Implementasi Sistem Deteksi Anomali Pergerakan Ayam Berbasis Web |
| Authors: | Wahjuni, Sri Akbar, Auriza Rahmad Henry, Matthew Martianus |
| Issue Date: | 2023 |
| Publisher: | IPB University |
| Abstract: | Dalam usaha ternak ayam broiler, kesejahteraan hewan merupakan salah satu aspek yang perlu diperhatikan peternak. Aspek dalam kesejahteraan hewan diantaranya adalah freedom from hunger or thirst (kebebasan dari kelaparan dan haus) dan freedom from pain, injury or disease (kebebasan dari rasa sakit, luka dan penyakit). Proses pemberian pakan secara terus menerus mengakibatkan meningkatnya massa ayam broiler, membuat ayam broiler menjadi kesulitan bergerak untuk mengambil makanan maupun air, yang mengakibatkan ayam broiler menjadi sakit kemudian mati. Tersedianya sistem yang dapat mendeteksi kesulitan bergerak pada ayam broiler akan memudahkan peternak dalam memantau pergerakan ayam broiler dari kejauhan. Dalam penelitian ini, deteksi ayam yang kesulitan bergerak dilakukan melalui integrasi model deteksi objek MobileNet-SSD dan algoritma deteksi anomali isolation forest dari penelitian sebelumnya. Integrasi dilakukan menggunakan bahasa pemrograman Python, diuji menggunakan video dengan kombinasi resolusi dan frame per second (FPS) tertentu, kemudian diimplementasikan menjadi sebuah aplikasi web menggunakan framework Flask. Hasil analisis integrasi menunjukkan bahwa semakin besar resolusi dan FPS video masukan, maka waktu inferensi dan waktu eksekusi dari sistem hasil integrasi menjadi semakin besar dan diestimasi meningkat secara linear. In the broiler farming business, animal welfare is an aspect that needs to be considered by breeders. Aspects of animal welfare include freedom from hunger or thirst and freedom from pain, injury, or disease. The process of continuous feeding results in broiler chicken's mass increase, making it difficult for them to move to get food and water, resulting in their sickness and death. The availability of a system that can detect the movement difficulty of the broiler chickens will make it easier for farmers to monitor the movement of broiler chickens from afar. In this study, the detection of chickens that had difficulty moving was carried out by integrating the MobileNet-SSD object detection model and the isolation forest anomaly detection algorithm from previous studies. Integration is carried out using Python programming language, tested using video with a combination of resolution and frame per second (FPS), then implemented into a web application using the Flask framework. The integration analysis shows that the greater the resolution and FPS of the input video, the greater the inference and execution time. Both times are also estimated to increase linearly. |
| URI: | http://repository.ipb.ac.id/handle/123456789/120101 |
| Appears in Collections: | UT - Computer Science |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| Cover.pdf Restricted Access | Cover | 294.28 kB | Adobe PDF | View/Open |
| G64190072_Matthew Martianus Henry.pdf Restricted Access | Fullteks | 957.02 kB | Adobe PDF | View/Open |
| Lampiran.pdf Restricted Access | Lampiran | 108.24 kB | Adobe PDF | View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.