1 Section of Chemical Engineering, The Faculty of Engineering and Science, Aalborg University, VBN2 Department of Chemistry and Bioscience, The Faculty of Engineering and Science, Aalborg University, VBN3 The Faculty of Engineering and Science (ENG), Aalborg University, VBN4 Samara State Technical University5 J&M Analytik AG6 Samara State Technical University
The applicability of conventional digital imaging to quantitative determination of fat and total protein in cow’s milk, based on the phenomenon of light scatter, has been proved. A new algorithm for extracting features from digital images of milk samples has been developed. The algorithm takes into account spatial distribution of light, diffusely transmitted through a sample. The proposed method has been tested on two sample sets prepared from industrial raw milk standards, with variable fat and protein content. Partial Least-Squares (PLS) regression on the features calculated from images of monochromatically illuminated milk samples resulted in models with high prediction performance when analysed the sets separately (best models with cross-validated R2=0.974 for protein and R2=0.973 for fat content). However when analysed the sets jointly the obtained results were significantly worse (best models with cross-validated R2=0.890 for fat content and R2=0.720 for protein content). The results have been compared with previously published Vis/SW-NIR spectroscopic study of similar samples.
Talanta: the International Journal of Pure and Applied Analytical Chemistry, 2014, Vol 121, p. 144-152