2014 – 2015 Coordination of Meat Science and carcase grading – Lamb and Beef

Summary

This report summaries co-ordination activities associated with the management of R&D and adoption of Meat Standards Australia and the Lamb Supply Chain Group.
Significant achievements include contributions to:
Transparent international collaboration on the science of MSA with Europe and ChinaPrototype cuts based MSA model for lamb developed which has emphasized the need for more sophisticated carcase and meat quality grading.Development of probe (mechanical, impedance and NIR) and hyperspectral imaging based measurement of lean meat yield and meat quality for beef and lamb carcases in collaboration with Carometec, MLA, MLA donor company and the Sheep CRCKey awareness and adoption activities for the development of lean meat yield and meat quality measurement in lamb with key processors and retailers.Maintained a strong relationship between the Sheep CRC/Lamb Supply Chain Group and MLA.Development of future R&D directions and plans for lean meat yield and eating quality in lamb and sheepmeatsAdoption and awareness activities for beef and lamb producers
Outcomes:
P.PSH.0631 Proof of concept for new lamb grading tools (Hyperspectrial imaging)
This project was completed to milestone 6, demonstrating proof of concept. The hyperspectral imaging of lamb cuts shows great promise segmenting tissue types, easing the task of determining loin eye area3D information is needed to avoid operator influence on area measurements3D information will provide the basis for edge detection, enabling accurate back fat measurementsInhomogeneous lighting and dynamic range of the camera greatly affect IMF accuracy and must be improvedImage acquisition speed must be improved to avoid displacement of individual images
It is very plausible to build a handheld grading system at a  reasonable cost that can accurately grade for loin eye area, subcutaneous fat depth and score IMF with an RMSEP around 1.5-2% against a chemical IMF reference along with colour of both fat and meat – the latter was not part of this project, but will be a requirement for cattle.
The hyperspectral instrument build for in this project is a first prototype. We believe, that with better wavelength selection, control of background light and 3D information using a small laser diode for instance, it will be possible to improve the scores as this should reduce the effects of operator handling.