Automatic estimation of meat characteristics using computer vision

Co-advisors:
Dan Tulpan, Animal Biosciences
Benjamin Bohrer, Food Science


Meat quality measurements and estimations are complex tasks with very high economic impact in the meat industry. Most of these measurements are currently performed manually for meat cuts, they are very time-consuming, inconsistent and the meat samples are destroyed in the process. It is therefore very important to develop automatic quality parameter estimations in post-mortem harvested meat. Ultrasound and color images are typically used in live body condition scoring and carcass estimations of meat tenderness and amount of fat, but to date, automatic estimation of meat quality is still in its infancy. This project aims to explore various computer vision approaches to automate the estimation process and will eventually lead to the development of a software prototype to achieve this task.