Early detection of sick chicken using artificial intelligence

Abstract

According to survey of BPS, in 2020 meat consumption especially poultry has reached 4 million tons per year or equivalent to 35% of total meal consumption. Every year the production and productivity of meat especially poultry has been increasing significantly. However, with such a large production volume, most of the process in poultry industry such as feeding, quality inspection, vaccine injection and sick detection in Indonesia still performed with human intervention. This process, like other manual activities, is very labor intensive, time consuming, and prone to inconsistencies and inaccuracies. This research is focus on early sick detection in poultry industry especially chicken farm. A delay in the sick detection stage can result in high economic losses for farmer. The automation process using computer vision technology is expected to reduce costs, increase efficiency and accuracy by applying the YOLOv5 algorithm. Based on the results of the study, conducted to 4333 feces images it was found that at the testing stage, the best average level of accuracy obtained was 89.2%.