Share Email Print
cover

Proceedings Paper

Nondestructive detection of pork quality based on dual-band VIS/NIR spectroscopy
Author(s): Wenxiu Wang; Yankun Peng; Yongyu Li; Xiuying Tang; Yuanyuan Liu
Format Member Price Non-Member Price
PDF $14.40 $18.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

With the continuous development of living standards and the relative change of dietary structure, consumers’ rising and persistent demand for better quality of meat is emphasized. Colour, pH value, and cooking loss are important quality attributes when evaluating meat. To realize nondestructive detection of multi-parameter of meat quality simultaneously is popular in production and processing of meat and meat products. The objectives of this research were to compare the effectiveness of two bands for rapid nondestructive and simultaneous detection of pork quality attributes. Reflectance spectra of 60 chilled pork samples were collected from a dual-band visible/near-infrared spectroscopy system which covered 350-1100 nm and 1000-2600 nm. Then colour, pH value and cooking loss were determined by standard methods as reference values. Standard normal variables transform (SNVT) was employed to eliminate the spectral noise. A spectrum connection method was put forward for effective integration of the dual-band spectrum to make full use of the whole efficient information. Partial least squares regression (PLSR) and Principal component analysis (PCA) were applied to establish prediction models using based on single-band spectrum and dual-band spectrum, respectively. The experimental results showed that the PLSR model based on dual-band spectral information was superior to the models based on single band spectral information with lower root means quare error (RMSE) and higher accuracy. The PLSR model based on dual-band (use the overlapping part of first band) yielded the best prediction result with correlation coefficient of validation (Rv) of 0.9469, 0.9495, 0.9180, 0.9054 and 0.8789 for L*, a*, b*, pH value and cooking loss, respectively. This mainly because dual-band spectrum can provide sufficient and comprehensive information which reflected the quality attributes. Data fusion from dual-band spectrum could significantly improve pork quality parameters prediction performance. The research also indicated that multi-band spectral information fusion has potential to comprehensively evaluate other quality and safety attributes of pork.

Paper Details

Date Published: 13 May 2015
PDF: 8 pages
Proc. SPIE 9488, Sensing for Agriculture and Food Quality and Safety VII, 94880T (13 May 2015); doi: 10.1117/12.2176831
Show Author Affiliations
Wenxiu Wang, China Agricultural Univ. (China)
Yankun Peng, China Agricultural Univ. (China)
Yongyu Li, China Agricultural Univ. (China)
Xiuying Tang, China Agricultural Univ. (China)
Yuanyuan Liu, China Agricultural Univ. (China)


Published in SPIE Proceedings Vol. 9488:
Sensing for Agriculture and Food Quality and Safety VII
Moon S. Kim; Kuanglin Chao; Bryan A. Chin, Editor(s)

© SPIE. Terms of Use
Back to Top