Baichen Yu 1 and Tyler Boulom 2, 1 USA, 2 Woodbury University, USA

This project addresses ecological and sanitation challenges associated with traditional backyard bird feeders by developing Ornimetrics, an intelligent, AI-enabled feeding system. Using an embedded camera and YOLOv11-based species recognition, the feeder identifies visiting birds, regulates food portions, and logs activity to a cloud database. The system reduces disease transmission risks, minimizes waste, and improves user awareness through real-time analytics. Two experiments were conducted: one assessing model accuracy across four species and another measuring species-specific waste patterns. Results showed strong detection accuracy with predictable weaknesses in low-light and high-motion conditions, as well as meaningful differences in waste generation. Methodology comparisons highlighted how Ornimetrics improves upon existing wildlife-monitoring and scene-classification frameworks by integrating automated decision-making into its workflow. Overall, the system demonstrates a viable and innovative approach to responsible, data-driven bird feeding.

Wildlife Monitoring, Computer Vision, Smart Feeding Systems, Ecological Sustainability