作者: Ahmed Rady , Akinbode Adedeji
DOI: 10.1016/J.MEATSCI.2017.10.014
关键词:
摘要: The main objective of this study was to investigate the use spectroscopic systems in range 400-1000nm (visible/near-infrared or Vis-NIR) and 900-1700nm (NIR) assess estimate plant animal proteins as potential adulterants minced beef pork. Multiple machine learning techniques were used for classification, adulterant prediction, wavelength selection. Samples first evaluated presence absence (6 classes), secondly type classes) level. Selected wavelengths models generally resulted better classification prediction outputs than full wavelengths. stage rates 96% 100% pure/unadulterated adulterated samples, respectively. Whereas, second had 69-100%. optimal predicting levels yielded correlation coefficient, r 0.78-0.86 ratio performance deviation, RPD, 1.19-1.98. results from illustrate application technology rapidly accurately detect