Prediction of moisture content in steamed and dried purple sweet potato using hyperspectral imaging analysis

Abstract

Partial least squares regression (PLSR) modeling was performed to predict the moisture content in steamed, dried purple sweet potato based on spectral data obtained from hyperspectral imaging analysis. The PLSR model with a combination of multiplicative scatter correction, Savitzky-Golay, and first derivative exhibited the highest accuracy (R P 2 = 0.9754). The wavelengths found that strongly affected the PLSR model were 961.12, 1065.50, 1083.93, 1173.23, and 1233.89 nm. These wavelengths were associated with the O-H second overtone and the second overtone of C-H, C-H 2 , and C-H 3 . When PLSR modeling was performed using these selected wavelengths, the prediction accuracy of the PLSR model exhibited high accuracy (R P 2 = 0.9521). Therefore, the moisture content could be predicted with high accuracy using only five wavelengths rather than the full spectrum.

NIR sensors combined with chemometric algorithms in intelligent quality evaluation of sweetpotato roots from ‘Farm’ to ‘Table’: Progresses, challenges, trends, and prospects

“…The selected five wavelengths improved the prediction efficiency while reduced the accuracy obviously (R 2 P = 0.913, RMSEP = 0.058) ( Su et al, 2019 ). During drying process, the moisture was quantified with excellent performance observed by applying PLS and BPANN (R 2 P > 0.95) based on either full range spectra or selected optimal wavelengths ( Heo, Choi, Kim, & Moon, 2021 ; Su et al, 2020 ), significantly better than using MLR and SVM algorithm ( Peng et al, 2021 ; Sun et al, 2017 ). It was indicated that LWNIR spectra was more suitable for moisture prediction than SWNIR spectra, which may due to the presence of more functional groups related to water absorption in the LWNIR region.…”

“…It was indicated that LWNIR spectra was more suitable for moisture prediction than SWNIR spectra, which may due to the presence of more functional groups related to water absorption in the LWNIR region. According to the dryness degree, the purple-flesh sweetpotato were discriminated with overall accuracy of >80% ( Heo et al, 2021 ). In addition, considering SWNIR data (400–1050 nm), the anthocyanin content in purple-flesh sweetpotato slices during drying process was estimated with satisfying results (0.86 < R 2 P < 0.90, RMSEP≦0.302 mg/g) ( Liu et al, 2017 ; Peng et al, 2021 ), allowing for the slight better performance from SVM prediction than from MLR calculation, which meant that there may be more nonlinear relationships between NIR data and anthocyanin concentration than linear ones.…”

Non-Destructive In-Process Quality Evaluation Of Plant-Sourced Food During Drying

“…Heo et al [31] steamed and dried purple sweet potato at 55 °C at different drying times of 0, 2, 4, 6, 8, or 10 h. They utilized the partial least squares regression (PLSR) model to project the moisture content based on hyperspectral imaging data analysis. The investigator found that the model utilised five wavelengths, which affected the PLSR model the most, provided a more accurate prediction than that of the model that used the full spectrum wavelength.…”

Detection of rice (with husk) moisture content based on hyperspectral imaging technology combined with MSLPP–ESMA–SVR model

“…Sun et al (2020) used 417–1000 nm HSI technology to non‐destructively determine the moisture content of barley seeds, and the prediction result showed that the determine coefficient reached 0.883 and the root mean square error reached 0.0198%. Suhyeon et al (2021) selected five feature wavelengths based on spectral data obtained from HSI analysis, and established PLSR model to predict the moisture content of evaporated‐dried purple potato, with a prediction accuracy rate of 0.9521. However, research on detecting moisture content in rice using HSI technology has been limited.…”