Technical note: Estimating body weight and body composition of beef cattle trough digital image analysis

Area of Science:

• Animal Science
• Agricultural Technology
• Biotechnology

Background:

• Traditional methods for assessing beef cattle body weight (BW), hot carcass weight (HCW), and body composition are labor-intensive and costly.
• Digital imaging presents a potential non-invasive, efficient alternative for livestock evaluation.
• Developing accurate predictive models is crucial for optimizing beef production and management.

Purpose of the Study:

• To develop predictive equations for body weight (BW), hot carcass weight (HCW), and empty body fat (EBF) in young bulls using digital images.
• To evaluate the efficacy of measurements derived from Microsoft Kinect depth videos for estimating key beef cattle metrics.
• To identify the most significant body measurements and derived indexes for predicting cattle weight and body fat.

Main Methods:

• Utilized a Microsoft Kinect device to capture infrared depth videos of 35 young bulls (Angus and Nellore) before slaughter.
• Extracted key body measurements (e.g., chest width, body length) from selected video frames.
• Generated 23 indexes from body measurements and performed regression analysis to develop predictive models for BW, HCW, and EBF.

Main Results:

• Chest width showed a strong correlation (r > 0.85) with BW and HCW.
• The predictive models for BW and HCW achieved R-squared values between 0.69 and 0.84.
• Index I5 (chest width × body length) significantly predicted BW and HCW, explaining over 70% of the variation.

Conclusions:

• Digital imaging with Microsoft Kinect offers a viable, cost-effective tool for estimating beef cattle BW and HCW.
• The developed predictive equations demonstrate the potential of image-based assessments in livestock management.
• Further research can refine these models for broader application in the beef industry.