Real-time augmentation of USDA yield grade application to beef carcasses using video image analysis

Area of Science:

• Agricultural Engineering
• Animal Science
• Food Science

Background:

• Accurate beef carcass yield grading (YG) is crucial for the meat industry.
• Traditional YG methods can be subjective and time-consuming.
• Video image analysis (VIA) offers potential for objective and efficient grading.

Purpose of the Study:

• To assess the ability of VIASCAN and Computer Vision System (CVS) VIA instruments to augment beef carcass YG.
• To evaluate the cutout prediction accuracy of a YG augmentation system at commercial packing plant speeds.
• To determine if VIA systems can operate in real-time for commercial application.

Main Methods:

• Two-phase study involving 505 (Phase I) and 290 (Phase II) beef carcasses.
• On-line USDA graders assigned YG and Quality Grades (QG).
• VIASCAN and CVS measured longissimus muscle area (LMA); kidney, pelvic, and heart fat (KPH) was weighed.
• Augmented YG were computed using instrument-measured LMA and KPH data.
• Comparison of augmented YG with expert YG and fabricated subprimal yields.

Main Results:

• Instrument-measured LMA showed high correlation with expert LMA (r = 0.88-0.94).
• Augmented YG using VIA instruments were closely related to expert YG (r = 0.92-0.95).
• Augmented YG accounted for 55-71% of the variation in fabricated subprimal yields.
• VIA systems demonstrated capability to operate at commercial plant speeds (340 carcasses/hour).

Conclusions:

• VIASCAN and CVS effectively augment beef carcass YG assignment.
• VIA systems can operate at commercial speeds, improving grading accuracy and efficiency.
• The technology shows promise for enhancing official USDA YG application in real-time.