Hyperspectral imaging in tandem with multivariate analysis and image processing for non-invasive detection and visualization of pork adulteration in minced beef

Mohammed Kamruzzaman, Yoshio Makino, Seiichi Oshita

Research output: Contribution to journalArticlepeer-review

Abstract

Pork adulteration in minced beef was detected for the first time using a hyperspectral imaging (HIS) technique. Minced beef samples were adulterated with minced pork in the range of 2-50% (w/w) at approximately 2% intervals. Images were acquired using a visible near-infrared hyperspectral imaging (VNIR-HSI) system and their spectral data were extracted. Several data pre-treatments and different linear multivariate analyses, namely partial least squares regression (PLSR), principal component regression (PCR), and multiple linear regression (MLR), were investigated to determine the predictive ability of VNIR-HSI in detecting pork meat adulteration in minced beef. PLSR had a better performance than PCR for predicting pork adulteration in minced beef. Only four wavelengths centered at 430, 605, 665, and 705 nm were selected as the important wavelengths to build the MLR model for visualizing the distribution of adulteration. The results confirm that HSI can be used to provide a rapid, low cost, and nondestructive testing technique for detection of adulteration in minced meat.

Original languageEnglish (US)
Pages (from-to)7496-7502
Number of pages7
JournalAnalytical Methods
Volume7
Issue number18
DOIs
StatePublished - Sep 21 2015
Externally publishedYes

ASJC Scopus subject areas

  • Analytical Chemistry
  • Chemical Engineering(all)
  • Engineering(all)

Fingerprint

Dive into the research topics of 'Hyperspectral imaging in tandem with multivariate analysis and image processing for non-invasive detection and visualization of pork adulteration in minced beef'. Together they form a unique fingerprint.

Cite this