Chicken pen is a very important component in a broiler farm. Therefore, the conditions in the cage and the condition of the chickens must also be maintained by the breeder. Conditions in a comfortable chicken pen will certainly affect the health and quality of the chickens produced. A simple way to see the health condition of farm chickens is to observe their diet. However, an inspection of the health condition of broilers is still carried out by manual inspection in commercial rearing houses. This manual process requires a lot of time and labor. With the development of the times, these inspections can be carried out without human intervention in the cage. One solution is to use Object Detection. Object detection is one of the most popular fields in computer vision and artificial intelligence (AI). In this study, one of the algorithms of object detection is used, namely YOLOv4. The algorithm is an algorithm developed to detect an object in real-time. The algorithm training process is carried out using 1,029 broiler images divided into eight classes. The testing process is carried out on videos captured in real-time by CCTV. Based on the results of the study, the best accuracy rate was 79.54% with a processing time of 7 seconds to be able to process 1,029 digital images.Keywords: Broiler, Computer Vision, YOLOv4
Copyrights © 2022