Value based trading system: image analysis of sheep and beef carcasses- proof of concept
| Project start date: | 28 February 2014 |
| Project end date: | 19 January 2015 |
| Publication date: | 01 January 2015 |
| Project status: | Completed |
| Livestock species: | Grassfed cattle, Grainfed cattle |
| Relevant regions: | National |
|
Download Report
(0.7 MB)
|
|
Summary
This project has provided a 'proof of concept' for estimating lean meat yield (kg) and consequently retail beef yield in beef carcasses. The position of the RGBD cameras has been determined for the scanning of beef and sheep carcasses. Techniques presented in this report are also transferable to estimating lean meat yield and consequently retail yield in sheep carcasses. The curvature of the butt profile along with hot carcass weight produced a strong relationship between lean mean yield and the signature profile of the carcasses in 2 datasets; 2.97kg and 5.03kg root mean square error (RMSE) in estimating lean meat yield in cattle carcasses in slaughters conducted in Feb 2014 (n=32) and May 2014 (n=31), respectively. However, on this small dataset, muscle score and hot carcass weight (HCW) estimated lean meat yield marginally better than curvature and HCW. This result highlights the important role that muscle score plays in estimating phenotypic traits in beef cattle.
More information
| Project manager: | Alex Ball |
| Primary researcher: | NSW Department of Industry |
