Using computer vision technology to evaluate the meat tenderness of grazing beef

Abstract

Raw meat surface features from non-grazing animals are reported to be correlated with meat tenderness. However, meat from grazing beef may have different tenderness to that of non-grazing beef due to differences in activity and diet. The feasibility of using meat surface characteristics from grazing beef in New Zealand to estimate meat sensory tenderness was tested. Results from striploin samples from 50 carcasses demonstrated that geometric, spectral and textural characteristics of meat from grazing beef were correlated to meat tenderness assessed by trained tasting panels. Correlations were obtained using a neural network approach (adjusted R2 = 0.62) and a linear multivariable regression technique (adjusted R2 = 0.58).