Rapid determination of protein, starch and moisture content in wheat flour by near-infrared hyperspectral imaging

Simultaneous Detection of Amylose and Amylopectin Content of Foxtail Millet Flour with Hyperspectral Imaging

“…In the past, scholars generally used hyperspectral imaging for the classification of foxtail millet varieties and the detection of nutritional components [40][41][42]. This study combined hyperspectral imaging with chemometrics to set different rates of sheep manure application and collected 358 samples of foxtail millet flour for the detection of amylose and amylopectin content.…”

Advancements, limitations and challenges in hyperspectral imaging for comprehensive assessment of wheat quality: An up-to-date review

“… Caporaso et al (2018) applied HIT in 980–2500 nm to determine protein distribution in whole single wheat kernels, obtaining prediction models with R 2 P more than 0.8 and an error below 1 %, indicating an application potential in breeding wheat to select kernels based on their protein content. With two narrower spectral bands (1120–2424 nm, 969–2174 nm), excellent performance in predicting protein content in wheat flour was found by Morales-Sillero et al (2018) (R 2 P = 0.99, RMSEP = 0.21 %) and Zhang et al (2023) (R 2 P = 0.9859, RMSEP = 1.1580 g/100 g), respectively, which further highlighted the great potential of HIT in determining protein in a fast and non-destructive way. Moisture and starch contents of wheat flour were also measured and visualized by HIT using 28 and 52 effective wavelengths selected from the 969–2174 nm range, respectively, exhibiting a better effect of predicting starch (R 2 P > 0.9) than predicting moisture (R 2 P < 0.9)( Zhang et al, 2023 ).…”

“…With two narrower spectral bands (1120–2424 nm, 969–2174 nm), excellent performance in predicting protein content in wheat flour was found by Morales-Sillero et al (2018) (R 2 P = 0.99, RMSEP = 0.21 %) and Zhang et al (2023) (R 2 P = 0.9859, RMSEP = 1.1580 g/100 g), respectively, which further highlighted the great potential of HIT in determining protein in a fast and non-destructive way. Moisture and starch contents of wheat flour were also measured and visualized by HIT using 28 and 52 effective wavelengths selected from the 969–2174 nm range, respectively, exhibiting a better effect of predicting starch (R 2 P > 0.9) than predicting moisture (R 2 P < 0.9)( Zhang et al, 2023 ). With 16 feature wavelengths selected from hyperspectral data (968–2576 nm) by successive projections algorithm (SPA), combined with Relieff-selected terahertz feature data, the ash content in wheat flour was well-predicted by a non-linear hierarchical extreme learning machine (H-ELM) model (R 2 P = 0.989, RMSEP = 0.015 %) proposed by Li et al (2023) .…”

Intelligent System/Equipment for Quality Deterioration Detection of Fresh Food: Recent Advances and Application

“…Images are usually represented mathematically as matrices. As shown in Formula (11) [ 97 ], the corrected image is obtained after subtraction and inverse matrix operations. where R is the corrected image, I is the original spectral image, W is the whiteboard data, and D is the blackboard data.…”