Segmentation of beef marbling based on vision threshold

Accurate segmentation of beef-marbling images plays an important role in making the correct decision on the beef-marbling score in an automatic beef quality grading system. The purpose of this study is to develop a novel segmentation method to correctly separate the fat flecks from the muscle in the rib-eye region in a beef image and to compare the proposed segmentation method to prior algorithms. The authors first examine the vision threshold of graders using visual test cards designed by authors, and then propose a segmentation method of the beef-marbling image based on the grader's vision thresholds. Finally, the proposed method and eight: other methods of image segmentation are applied to the segmentation of the sample images and an assessment of algorithm performances was made by a five-member panel on the basis of visual inspection. The results show that the proposed approach performs most robustly in comparison with the others. (c) 2008 Elsevier B.V. All rights reserved.