Application of computer vision and artificial intelligence to monitor poultry behavior under various stress factors

УДК 636.5.033
DOI 10.33861/2071-8020-2024-6-41-45

Zelenkov A. P., Fomina A. S., Kochetkova N. A., Vasiliev P. V.

Summary. The study presents an analysis of modern publications devoted to the identification of signs of problematic behavior of agricultural and wild birds under conditions of technological stress. The main attention is paid to factors that cause stress in poultry, such as intensive farming methods, changes in habitat and anthropogenic loads. Key behavioral markers of stress are indicated, such as excessive pinching, changes in vocalization, and abnormal social interactions. Common manifestations of problematic behavior are considered: aggression, decreased appetite, impaired reproductive functions and anomalies in migration patterns. Adaptation mechanisms and possible stress reduction strategies are discussed, including improved living conditions and access to natural activities. technological stress. The necessity of rethinking approaches to bird health management using computer vision systems is substantiated. The integration of new technologies for monitoring the condition of birds will allow timely identification and correction of sources of stress. This determines the prospects for the introduction of computer vision systems in production. Despite the high technological complexity of such developments, the uniqueness of the resulting datasets, the need for individual training and retraining of neural networks for each task, their use remains promising for early detection of small behavioral signs of bird distress during dense planting. The need for further research in this area is emphasized in order to develop a more humane approach to the maintenance of poultry. Recommendations are also offered for improving management practices, which can contribute to improving the welfare of birds and increasing the sustainability of their populations in the face of increasing technological pressure.

Keywords: technological stress, heat stress, agricultural poultry, wild poultry, computer vision, artificial intelligence, precision animal husbandry, problem bird behavior.

Author affiliation:

Fomina Anna S., Ph. D. in Biology, Docent, Docent of the Department of Biology and General Pathology of the Don State Technical University; 1, Gagarina sq., Rostov-on-Don, 344003; phone: 8-908-1742627; e-mail: a_bogun@mail.ru.

Kochetkova Nataliya A., Senior Lecturer of the Department of Biology and General Pathology of the Don State Technical University; 1, Gagarina sq., Rostov-on-Don, 344003; phone: 8-988-5569758; e-mail: les.nata13@bk.ru.

Vasiliev Pavel V., Ph. D. of Technics, Docent of the Department of Information Technology of the Don State Technical University; 1, Gagarina sq., Rostov-on-Don, 344003; phone: 8-918-5036280; e-mail: lyftzeigen@mail.ru.

Responsible for correspondence with the editorial board: Zelenkov Alexey P., D. Sc. in Agriculture, Professor of the Department of Biology and General Pathology of the Don State Technical University; 1, Gagarina sq., Rostov-on-Don, 344003; phone: 8-951-8176966; e-mail: zelenkovalex@rambler.ru.