We consider the use of optical coherence tomography (OCT) imaging to predict the quality of meat. We find that intramuscular fat (IMF) absorbs infrared light about nine times stronger than muscle, which enables us to estimate fat content in intact meat samples. The method is made very efficient by extracting relevant information from the three-dimensional high-resolution images generated by OCT using principal component analysis (PCA). The principal components are then used as regressors into a support vector regression (SVR) prediction model. The SVR model is found to predict IMF content stably and accurately, with an R2 value of 0.94. Our study paves the way for automated, contact-less, non-destructive, real time classification of the quality of meat samples.
Keywords: Machine learning; Meat quality; Optical coherence tomography; Principal component analysis; Support vector regression.
Copyright © 2021 The Authors. Published by Elsevier Ltd.. All rights reserved.