A Digital Image Method for Measuring and Analyzing Color Characteristics of Various Color Scores of Beef

A digital image method was developed to quantitatively describe and analyze color features of beef lean with various grades of color and to study the relationships between characteristics and scores of beef color. A lighting system was specially designed and constructed for image acquisition. One hundred and sixty of representative beef samples were selected from a local supplier and used for color measurement and sensory evaluation. After segmentation, nine of features characterizing color of the entire longissimus dorsi (l.d.) muscle were extracted from the segmented images and analyzed. Furthermore, correlation analyses between these color features and color grade of samples were carried out. Description statistics showed that there was a narrow range of color feature values for beef lean purchased in market places in China. Significant distinctions in lean color features among different color score of beef were confirmed (P<0.001). Color characteristics of mean R, G, B, H, S, I, L^* and b^* values except for a value were significant correlated (p<0.01) with color scores of beef. Image processing was an effective method for measuring and analyzing color attributes of fresh beef lean.