A feasibility quantitative analysis of NIR spectroscopy coupled Si-PLS to predict coco-peat available nitrogen from rapid measurements

Lu, Bing; Liu, Nihong; Wang, Xufeng; Hu, Can; Tang, Xiuying

doi:10.1016/j.compag.2020.105410

Cited by 16 publications

(8 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…(S‐G)). S‐G smoothing is used to smooth spectral data, and the window size will directly affect the final smoothing effect (Lu et al., 2020). SNV + D is a combination of SNV pretreatment and detrending pretreatment.…”

Section: Methodsmentioning

confidence: 99%

“…The computation time can be long depending on the number of intervals and the selected number of intervals to combine. Finally, the combined local intervals that presented the lowest error values were selected (Lu et al., 2020; Sampaio et al., 2017).…”

Section: Methodsmentioning

confidence: 99%

“…Near‐infrared spectroscopy (NIRS) represents the vibration response of different molecules inside the sample in the near‐infrared (NIR) region (Chu, 2011), and it mainly reflects the absorption of molecular overtone and combination vibrations of hydrogenous groups X ‐ H (X = C, N, O) of the sample. As a fast, convenient, and green analysis method, it has been widely used in various fields, including dairy products (Zhang et al, 2014), meat (Zheng et al, 2019), medicine (Pereira et al, 2016), petrochemicals (Paiva et al, 2015), agricultural production (Lu et al, 2020), and so forth. Based on the principle of NIRS, it is feasible and effective to use NIRS combined with multiple regression analysis to detect the content of FFAs in rice, but it has not been fully exploited.…”

Section: Introductionmentioning

confidence: 99%

“…In order to obtain more effective information, the researchers had proposed to use chemometric methods to improve the signal‐to‐noise ratio of the spectra data. For example, multiple scattering correction pretreatment is used to improve the scattering effect (Jiang et al., 2020), Savitzky–Golay (S‐G) smoothing is used to smooth the noise (Lu et al., 2020), derivative pretreatment is used to improve the signal‐to‐noise ratio of the spectrum (L. Li et al., 2019), and so forth. Except that the prediction accuracy of the prediction model can be further improved by selecting the key variables related to the concentration value.…”

Section: Introductionmentioning

confidence: 99%

“…For example, the competitive adaptive reweighted sampling (CARS) algorithm selects variables with higher regression coefficient and removes useless variables to improve the accuracy of the prediction model (H. Li et al., 2009). Synergy intervals partial least squares (Si‐PLS) algorithm divides the spectrum into several intervals and selects a certain amount of intervals from these intervals to recombine into a new local interval, and the local interval data with the minimum root‐mean‐square error is considered to be the key intervals (Lu et al., 2020; Sampaio et al., 2017). Although there are many mature chemometrics methods, none of them has been fully used in the detection of FFAs in rice by NIRS.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

A feasibility quantitative analysis of free fatty acids in polished rice by fourier transform near‐infrared spectroscopy and chemometrics

Liu

Peng

et al. 2021

Journal of Food Science

View full text Add to dashboard Cite

Free fatty acids (FFAs) are an important indicator of the freshness and quality of rice. In this study, the vibration response of C-H chemical bonds (-CH 3 , -CH 2 , H-C = C-H) of FFAs in the near-infrared region was determined by analyzing the standard reagent. In addition, the spectral data of different physical forms of rice and chemometrics, such as partial least squares (PLS), synergy interval-PLS, and competitive adaptive reweighted sampling (CARS), were applied to develop an optimal regression model for rice FFAs determination. The performance of the FFAs model established by using the polished rice granule spectrum (PRG) combined with CARS was the best, the correlation coefficients of the calibration set and prediction set were 0.99 (root mean squared errors of the calibration = 2.00 mg/100 g) and 0.98 (root mean squared errors of the prediction = 3.21 mg/100 g), respectively, and the ratio of performance-to-deviation was 4.50. Compared with the rice powder spectral, the PRG spectral can better retain the information of FFAs. The result shows that NIRS can rapidly, nondestructively, and accurately detect FFAs in rice granules, which will help rice business and food regulatory authorities to establish an early warning mechanism of rice aging.Practical Application: Free fatty acids (FFAs) in rice are an important indicator for evaluating the freshness of rice, and their high responsiveness to the deterioration of rice quality. The real-time detection of FFAs in rice can timely adjust the parameters of the rice storage environment, which is very meaningful to ensure the quality of rice.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%