Abstract
ABSTRACT Automating the carcass beef grading process requires that the visual capabilities of a human grader be simulated. One important requirement of a computer grading system is that it be able to locate the longissimus dorsi (l.d.) muscle in a scene containing an image of the exposed surface of the carcass' 12th rib. Characteristics of the l.d. are important in determining both the yield and quality grades. Previously developed image processing methods have been successful in segmenting the loin eye area, but the results tend to include extraneous muscle tissue not part of the l.d. This article presents two image processing algorithms for separating these non-l.d. components. Both are based on the assumption that the scene can be modeled as a union of overlapping convex sets, and both are implemented using mathematical morphology. The first algorithm is a modification of one proposed by Meyer (1979) to separate overlapping convex blobs. The second uses a morphological opening to destructively remove external portions. Both methods were found to perform adequately when tested on 30 images of beef carcass ribeyes.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
