Show simple item record

dc.contributor.advisorWahjuni, Sri
dc.contributor.advisorAkbar, Auriza Rahmad
dc.contributor.authorFathurrahman, Muhammad Ezra
dc.date.accessioned2023-08-24T02:25:45Z
dc.date.available2023-08-24T02:25:45Z
dc.date.issued2023
dc.identifier.urihttp://repository.ipb.ac.id/handle/123456789/124289
dc.description.abstractProduktivitas daging ayam merupakan hal yang penting dalam menjaga pasokan daging ayam sebagai komoditas penting dalam negeri. Pemantauan terhadap perilaku agresif ayam merupakan salah satu upaya menjaga kesejahteraan ayam. Penelitian untuk mengamati perilaku ayam telah dilakukan dengan bantuan deep learning yaitu YOLOV4 yang menghasilkan model deteksi perilaku agresif ayam. Penelitian ini bertujuan mengimplementasikan model tersebut dalam bentuk website dengan menggunakan framework Flask dengan library OpenCV. Pengguna dapat mengunggah video yang ingin dilakukan deteksi perilaku agresif lalu sistem akan melakukan deteksi dengan YOLOv4 dan menghasilkan video yang telah ditandai. Pengujian implementasi pada web dilakukan dengan memasukkan beberapa jenis video yang memiliki resolusi dan durasi berbeda. Hasil uji menunjukkan bahwa deteksi video berhasil dilakukan dan ditampilkan keterangan tambahan berupa average confidence dan durasi waktu proses. Dapat disimpulkan bahwa, semakin besar resolusi dan durasi video masukan maka semakin tinggi nilai confidence yang didapat dan semakin lama proses deteksinya.id
dc.description.abstractThe chicken meat productivity is important to maintain the chicken meat supply as an important domestic commodity. The monitoring of the chicken’s aggressive behavior is an effort to maintain their welfare. Many research to observe chicken behavior has been carried out with the help of deep learning, namely YOLOv4 which produces a model to detect the chicken’s aggressive behavior. This study aims to implement the model in the form of a website. It was created using the Flask framework with the OpenCV library. A user can upload video that he wants to detect aggressive behavior, then the system will detect them with YOLOv4 and produce video that have been tagged. Implementation testing on the web is carried out by entering several types of videos that have different resolutions and durations. The test results show that video detection was successful and additional information is displayed in the form of average confidence and processing time duration. In conclusion, the larger the resolution and duration of the input video, the higher the confidence value obtained and the longer the detection process.id
dc.language.isoidid
dc.publisherIPB Universityid
dc.titleImplementasi Model Yolov4 untuk Pemantauan Perilaku Agresif Ayam Broiler Berbasis Websiteid
dc.typeUndergraduate Thesisid
dc.subject.keywordAggressive Behaviorid
dc.subject.keywordBroiler Chickenid
dc.subject.keywordObject Detectionid
dc.subject.keywordFlaskid
dc.subject.keywordYOLOv4id


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record