PSIX-21 Computer vision system as a tool to predict intramuscular fat of pork whole loin and chop.

The objective of this study was to examine the potential of using computer vision system (CVS) as a tool to predict pork intramuscular fat (IMF) percentage under industry scale equipment and environment. In this project 200 pork loins and its anterior (3rd rib) and posterior (10th rib) chop were collected from 7 different packing plants (n=1400 loins; n=2800 chops). Color images of pork whole loins and individual chops were acquired using our own developed CVS. Images were then segmented to lean muscle pixels and IMF pixels to calculate image IMF%. Subjective marbling scores (SMS) were assigned on a scale from 1 to 10 according to the National Pork Board standards (NPB, 2011). Crude fat percentage (CF%) was calculated using ether extract method (AOAC, 1990). The average CF% of anterior and posterior chop was used to represent the CF% of whole loin. Software SAS (v. 9.4; SAS Institute, Inc., Cary, NC, USA) was adopted to estimate the simple statistics for CF%, SMS, and image IMF%. Results shown that SMS had an overall accuracy of 53.3% and 70.1 % for predicting CF% of whole loin and individual chops. Comparatively, the overall accuracy of using CVS was 58.56% and 68.6 % for whole loin and individual chops. While neither SMS and CVS exhibited satisfying accuracies in predicting CF%, these results demonstrate the potential of using CVS as an objective measurement for pork whole loins and individual chops CF% within the industry environment and even the potential of replacing subjective marbling scores in the future.