Yolov8-Based Early Detection Model for Lame Broilers

Lameness is leg disorder which is caused by vitamin deficiency in broilers. The prediction of broiler diseases at an early stage is important as it safe farmers from huge economic loss. Also, broiler diseases cause harm to human health. To predict various broiler diseases large amount of work has been done. However, most of the existing research focused on older broilers (30–45 days),but its time to work for younger broilers (1–15 days) for early detection. Additionally, only a few researchers considered broilers in a group for disease prediction. In a real-time setting, broilers usually sit in groups. So, there is a requirement for a model which can detect broilers in groups. Broiler videos are collected from poultry farms for this purpose. In this research, an early detection model is proposed for identifying lame and healthy broilers. To detect lame broilers, YOLOV8 is employed. YOLOv8 obtain a precision of 95.7%, recall of 96.8%, and mAP of 94.7%. This research is conducted under controlled scenarios for the measurement of lame and healthy broilers. To predict the lameness of broilers, the posture and behavior parameters establish a strong foundation for the prediction of individual and groups of lame broilers. The outcomes shows that the YOLOV8 model is capable for early prediction of lame broilers.