Please use this identifier to cite or link to this item: http://repository.ipb.ac.id/handle/123456789/153326
Title: Deteksi Parasit Theileria equi dan Babesia caballi Secara Cepat dan Akurat Berbasis Algoritma YOLO-v8
Other Titles: Fast and Accurate Detection of Theileria equi and Babesia caballi Parasite Based on Yolo-v8 Algorithm
Authors: Nugraha, Arifin Budiman
Arif, Ridi
Kedaton, Feni Gemala
Issue Date: 2024
Publisher: IPB University
Abstract: Piroplasmosis merupakan penyakit yang disebabkan oleh parasit Theileria equi dan Babesia caballi. Penyakit ini menyebabkan kerugian dalam industri kuda, diantaranya kuda mengalami anemia, ikterus, kegagalan organ, serta larangan keikutsertaan dalam kompetisi olahraga berkuda. Oleh karena itu, penelitian ini bertujuan membuat sistem identifikasi parasit T. equi dan B. caballi secara otomatis berdasarkan algoritma YOLO. Pembuatan sistem identifikasi dilakukan dengan melakukan pemotretan preparat ulas darah T. equi dan B. caballi yang diwarnai oleh pewarna Giemsa 10%, pembuatan dataset, anotasi dataset, pengembangan sistem, dan pengujian kemampuan sistem. Pengembangan sistem untuk mendeteksi stadium T. equi dan B. caballi dibantu oleh mitra penelitian PT. Vox Digital Kreatif. Sistem identifikasi berhasil dikembangkan dan menunjukkan kemampuannya dalam mendeteksi T. equi dan B. caballi dengan tingkat akurasi mAP50 69,8%, mAP50-95 40,5%, dan kecepatan deteksi 5,4 ms. Uji performa sistem secara manual mendapatkan nilai akurasi 91%, presisi 98%, recall 92%, dan F1 Score 95%. Hasil penelitian menunjukkan sistem berhasil mengidentifikasi parasit B. caballi dan T. equi secara cepat dengan presisi yang tinggi.
Piroplasmosis is a disease caused by the parasites Theileria equi and Babesia caballi. This disease causes losses in the horse industry, including horses experiencing anemia, jaundice, organ failure, and a ban on participation in equestrian sports competitions. Therefore, this study aims to create an automatic identification system for T. equi and B. caballi parasite based on YOLO algorithm. The creation of the identification system was carried out by photographing the blood smear of T. equi and B. caballi which is stained with 10% Giemsa dye, creating a dataset, dataset annotation, system development, and system capability testing. The development of the system to detect T. equi and B. caballi was assisted by research partner PT. Vox Digital Kreatif. The identification system was successfully developed and demonstrated its ability to detect T. equi and B. caballi with 69,8% mAP50 accuracy, 40,5% mAP50- 95, and 5,4 ms detection speed. Manual system performance tests obtained 91% accuracy, 98% precision, 92% recall, and 95% F1 Score. The results showed that the system was able to identify B. caballi and T. equi parasites quickly with high precision.
URI: http://repository.ipb.ac.id/handle/123456789/153326
Appears in Collections:UT - Veterinary Clinic Reproduction and Pathology



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.