Skip to main content
ARS Home » Southeast Area » Athens, Georgia » U.S. National Poultry Research Center » Quality and Safety Assessment Research Unit » Research » Publications at this Location » Publication #333151

Research Project: Assessment and Improvement of Poultry Meat, Egg, and Feed Quality

Location: Quality and Safety Assessment Research Unit

Title: Predicting color traits of intact broiler breast fillets using visible and near-infrared spectroscopy

item JIANG, HONGZHE - China Agricultural University
item Yoon, Seung-Chul
item Zhuang, Hong
item WANG, WEI - China Agricultural University

Submitted to: Journal of Food Analytical Methods
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 4/12/2017
Publication Date: 4/27/2017
Citation: Jiang, H., Yoon, S.C., Zhuang, H., Wang, W. 2017. Predicting color traits of intact broiler breast fillets using visible and infrared-light. Journal of Food Analytical Methods. 10:3443-3451.

Interpretive Summary: Color appearance of chicken breast fillets is an important quality attribute that directly affects the functional properties of the meat during further processing and the consumer acceptance of fresh products. Among various color models (called color spaces), the L*a*b* color space is a device-independent color space and often used to predict color appearance of chicken breast fillets. The objective of the study was to compare the performance of visible and near-infrared spectroscopy measured from either skin side or bone side of chicken fillets for predicting color traits (L*, a*, b*, Chroma, Hue angle) of fresh chicken breast meat and the effect of different spectral ranges on a prediction model. The study found that the prediction model based on spectra collected from the bone side performed better than those from the skin side in prediction of color traits of chicken breast fillets. The model trained with the spectra in the 400-1,100 nm (or 400-2,500 nm) spectral range performed better than those in the 1,100-2,500 nm and 900-1,700 nm ranges. These results suggest that it is feasible to accurately predict color appearance of fresh chicken breast meat with visible and near-infrared spectra collected from bone side in the wavelength range of 400-1,100 nm.

Technical Abstract: The ability of using visible and near-infrared (Vis/NIR) spectroscopy with wavelengths ranging from 400 to 2500nm to predict broiler breast fillets color traits was assessed in this study. Deboning fillets from 214 carcasses slaughtered in a commercial processing plant were included in this work, spectra of the skin sides and the bone sides were collected respectively. Prediction models were developed from the original dataset for different color indexes (L*, a*, b*, Chroma and Hue angle) using partial least squares regressions (PLSR). Data of different spectral ranges, including 400-2500nm, 400-1100nm, 1100-2500nm as well as 900-1700nm were tested individually. Predictive ability of the models was assessed by coefficient of determination of validation (Rv2), the ratio performance deviation (RPD) and the range error ratio (RER). Generally, color traits of bone sides were obviously better predicted than those of the skin sides, and the better results were obtained using spectra of 400-2500nm as well as 400-1100nm. Predictions were good for a* Rv2=0.81; RPD=2.21; RER=12.13) and hue angle (Rv2=0.80; RPD=2.07; RER=12.65), and there is considerable potential for L*, b* and chroma (Rv2=0.73, 0.72 and 0.72; RPD=1.94, 1.85 and 1.81; RER=8.94, 8.51 and 8.48, respectively). Results suggested that Vis/NIR spectroscopy might be a feasible method for the assessment of color traits of intact broiler breast fillets, especially when spectra of bone-side were adopted.