Innovations in Computer Vision for Sustainable Precision Livestock Management

Precision Livestock Farming (PLF) is increasingly recognized as a key enabler for improving sustainability, animal welfare, and production efficiency in modern livestock systems. Among emerging technologies, computer vision has gained significant attention due to its ability to provide non-invasive, continuous, and scalable monitoring of animals in real farming environments. Recent advances in deep learning, including convolutional neural networks, YOLO-based detectors, vision transformers, and zero-shot approaches, have substantially improved the performance of automated systems for animal detection, tracking, and behavior analysis. These developments are particularly relevant in dairy farming, where individual-level monitoring is essential for early disease detection, welfare assessment, and productivity optimization. However, real-world farm conditions still pose challenges such as occlusions, group housing complexity, and variability in environmental conditions. Addressing these issues is crucial to unlock the full potential of computer vision and to support the transition toward more sustainable and data-driven livestock management systems.This Research Topic aims to gather cutting-edge contributions that advance the application of computer vision in sustainable precision livestock management. While significant progress has been made in controlled environments, there remains a critical need for robust, scalable, and transferable solutions that can operate reliably under real farm conditions. Key challenges include accurate individual animal identification in open-set scenarios, resilience to occlusions and dynamic environments, and seamless integration with complementary technologies such as RFID and other sensor systems. We particularly encourage contributions that explore novel methodological frameworks, including metric learning, multimodal data fusion, and edge AI, to enhance the robustness and scalability of livestock monitoring systems. Submissions addressing cross-species generalization, modular system design, and deployment in commercial farming contexts are also highly relevant. By fostering interdisciplinary collaboration between computer vision researchers, animal scientists, and industry stakeholders, this Research Topic seeks to bridge the gap between technological innovation and practical adoption. Ultimately, the goal is to support the development of intelligent systems that improve animal welfare, optimize resource use, and contribute to global sustainability targets, including the United Nations Sustainable Development Goals.We welcome Original Research articles, Reviews, Methods papers, and Perspective articles that present novel methodologies, interdisciplinary approaches, or applied solutions with clear relevance to precision livestock farming. Contributions demonstrating validation in real farm environments are particularly encouraged. This Research Topic welcomes submissions that address, but are not limited to, the following themes:• Computer vision for animal detection, tracking, and identification in multi-animal environments• Open-set recognition and generalization across farms and species• Pose estimation and behavior analysis for health and welfare monitoring• Robust vision systems under occlusions and challenging environmental conditions• Integration of vision with RFID, IoT, and other sensing technologies• Edge computing and scalable AI solutions for real-time farm applications• Development of datasets, benchmarks, and evaluation methodologies for PLF• Case studies and real-world deployments in commercial livestock systems Topic editors: MARCO BOVO, Miel Hostens Source: https://www.frontiersin.org/research-topics/78535/innovations-in-computer-vision-for-sustainable-precision-livestock-management