Abstract: This study aimed to investigate color change of beef lean tissue during storage. Fresh beefs were stored at room and refrigeration temperatures, respectively. Their images were captured with a CCD camera every day. Then the computer image processing techniques were applied to segment the lean tissue region from each whole image and to extract color features of lean tissue image in RGB and HIS spaces. Statistic analysis was finally employed to quantitatively analyze changes of beef color during storage. It was found that beef color varies from luminous cerise to dark wine and, meanwhile, R value of beef lean tissue decreases linearly and G and B values rise linearly with storage duration extending. H value of beef color fluctuates in a small range and both H and I values linearly decrease as storage duration extending. Compared with human vision, computer vision method can quantitatively and more accurately reveal the change of beef color.

Key words: beef, color, machine vision