Click on “Download PDF” for the PDF version or on the title for the HTML version.


If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.

Broilers’ weight estimation through depth image analysis

Published by the American Society of Agricultural and Biological Engineers, St. Joseph, Michigan www.asabe.org

Citation:  2021 ASABE Annual International Virtual Meeting  2100803.(doi:10.13031/aim.202100803)
Authors:   Luana Maria Benicio, Késia O.S Miranda, Tami Brown Brandl, Joseph L. Purswelld, Sudhenu Raj Sharmab, Isabella C.F. S. Condottad
Keywords:   Depth sensor, image processing, real-time monitoring.

Abstract. A continuous weight monitoring system would help producers ensure broilers are growing as expected and help with scheduling slaughter. Traditionally, the average weight of a flock is estimated by visual assessment or by manually weighing a random sample of birds, which can be time-consuming and prone to errors. Searching for alternatives to the traditional method that are more efficient, faster, and less invasive becomes necessary. Weight prediction through image analysis is one such alternative. Therefore, this study‘s objective was to automatically obtain the body weight of broilers from body dimensions extracted from depth images. Ten depth images and weights of 80 birds of Cobb commercial line were collected at five different ages (8, 14, 21, 28, and 35 days old). Weights ranged from 125.5 g to 2558.15 g at 35 days old. The Kinect Azure® depth camera was used for image acquisition. Data analysis was performed using an algorithm developed on MATLAB® (R2018a). Body dimensions (minimum and maximum heights of standing and sitting birds, head to tail length and width between wings), projected body volume, and dorsal area were acquired and correlated with the measured weight using a multi-linear regression. Results indicate that broilers‘ body weight can be estimated from their body dimensions, projected volume, and dorsal area using a multi-linear regression model (R² = 0.94). Therefore, this study indicates that this model can be used as a tool to monitor broilers‘ body weight effectively, practically, and continuously during the production phase.

(Download PDF)    (Export to EndNotes)