Kang Mun-Hye, Oh Sang-Hyon
Division of Aerospace and Software Engineering, Gyeongsang National University, Jinju 52828, Korea.
Division of Animal Science, Gyeongsang National University, Jinju 52828, Korea.
J Anim Sci Technol. 2025 Jan;67(1):43-55. doi: 10.5187/jast.2025.e4. Epub 2025 Jan 31.
This review examines the application of video processing and convolutional neural network (CNN)-based deep learning for animal face recognition, identification, and re-identification. These technologies are essential for precision livestock farming, addressing challenges in production efficiency, animal welfare, and environmental impact. With advancements in computer technology, livestock monitoring systems have evolved into sensor-based contact methods and video-based non-contact methods. Recent developments in deep learning enable the continuous analysis of accumulated data, automating the monitoring of animal conditions. By integrating video processing with CNN-based deep learning, it is possible to estimate growth, identify individuals, and monitor behavior more effectively. These advancements enhance livestock management systems, leading to improved animal welfare, production outcomes, and sustainability in farming practices.
本综述探讨了视频处理和基于卷积神经网络（CNN）的深度学习在动物面部识别、鉴定和重新鉴定中的应用。这些技术对于精准畜牧业至关重要，有助于应对生产效率、动物福利和环境影响方面的挑战。随着计算机技术的进步，牲畜监测系统已发展为基于传感器的接触式方法和基于视频的非接触式方法。深度学习的最新进展能够对积累的数据进行持续分析，实现动物状况监测的自动化。通过将视频处理与基于CNN的深度学习相结合，可以更有效地估计动物生长、识别个体并监测行为。这些进展改进了牲畜管理系统，从而提高了动物福利、生产成果以及养殖实践的可持续性。
