Application of a computer vision system to classify beef as normal or dark, firm, and dry1

Abstract

The purpose of this research was to evaluate the possibility of using computer vision systems (CVS) to detect DFD beef. The experimental material consisted of beef obtained from 60 carcasses (M. semimembranosus) that were 48 h postmortem. The beef was classified into 2 quality groups according to pH: normal (pH < 5.8) or DFD (pH ≥ 5.8). L*, a*, and b* color components were determined in the experimental materials using both CIELab and a CVS using 3 color models: RGB (red, green, blue), HSV (hue, saturation, value), and HSL (hue, saturation, lightness). The total heme pigment content also was determined. As determined by both CIELab and CVS, DFD meat was characterized by a darker color than normal beef. Color lightness, as determined using both CIELab and CVS, can also be used to detect DFD beef. A significant relationship was found between V (r = -0.79) and L (r = -0.77) using CVS and pH, which is an indicator of DFD beef. This finding indicates the possibility of using these color components to assess the pH level of beef.