Location: Poultry Research
Project Number: 6064-32630-008-05-S
Project Type: Non-Assistance Cooperative Agreement
Start Date: Sep 1, 2020
End Date: Sep 30, 2021
The objective of this research to assess bird activity and behavior as related to management factors such as stocking density, lighting, feeding system, and ventilation conditions and correlate to production efficiency and characterize energy, water, and feed use requirements.
Behavior and locomotion will be measured using machine vision in research and commercial broiler houses using a combination of infrared, digital and depth imaging to determine locomotive, clustering, resting, posture changes, and feeding/drinking behaviors to assess their influence on metabolic responses including heat and moisture production. In parallel, environmental variables such as air temperature, globe temperature, humidity, air speed, lighting will be measured to correlate to bird behavior, posture, and production efficiency metrics such as average daily gain, feed efficiency, and mortality. Imaging platforms will be installed in both small- and large-scale floor pens to collect behavior data on multiple broiler flocks. Specific environmental stresses will be applied in the small-scale pens, and large-scale tests will be completed to ensure scalability. Performance data will be collected bi-weekly at each feed phase change for a 56-day rearing period. Yield data will be collected at a university pilot processing plant. Digital and depth cameras will be placed on the ceiling over the feed lines and in select resting areas in the pens. This camera orientation will be selected to ensure that maximum bird monitoring can occur with little or no inference with daily activities. Different image frequencies will be tested, from video (at 10 frames/sec) for a portion of the day a single image every 5 seconds. Image frequencies will be tested to find the slowest data collection rate that provides adequate data. Several algorithms will be developed. The initial focus of these research will include image processing to determine: 1) bird weight by calculating volume, 2) environment stress by detecting posture changes, and 3) social stress by determine changes in activity in the location of the feeder.