Scientists from Nanjing Forestry University recently used a portable hyperspectral imager to help detect maturity stages in Camellia oleifera fruits. Their findings were published in Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy (1).
Camellia oleifera fruits, which are extensively cultivated in various regions in southern China, are considered one of the four major woody oil crops globally, putting it alongside olives, palms, and coconuts. To obtain the highest amounts of seed yield and oil content, scientists must determine the optimal harvesting stage criteria for Camellia oleifera. This need has led to efforts by researchers to accurately identify the maturity stage of Camellia oleifera fruits.
Currently, this process is complex, and it is prone to being affected by different intrinsic and environmental factors. To rectify this problem, the scientists behind this study proposed a non-invasive detection method based around hyperspectral imaging (HSI) technology. HSI integrates spectral and imaging techniques to simultaneously record spectral and spatial information, with the approach being used most often in the food industry for the nondestructive assessment of food quality. For example, a recent LCGC International piece focused on how hyperspectral imaging could be used to gather data on bruises in Fuji apples, predicting when these imperfections were caused during the delivery process and how much the bruises affect the overall quality of the apples (2). As a result, HSI is the best method for assessing the internal quality of fruit and its maturity stage while allowing for the acquisition of comprehensive sample information and generating large amounts of spectral and spatial data.
Read More: Hyperspectral Images of Fuji Apples Used as Predictive Data for Fruit Bruise Area
In their study, the researchers used a portable hyperspectral imager for the in-field image acquisition of Camellia oleifera fruits at three maturity stages. Using this imager, 10 quality indexes were measured as reference standards. Next, factor analysis was performed to obtain the comprehensive maturity index (CMI); specifically, the scientists analyzed the change trends and correlations of different indexes. To lessen the high dimensionality of spectral data, a successive projection algorithm (SPA) was used to select effective feature wavelengths. Various CMI prediction models were constructed based on full spectra and feature wavelengths; these prediction models included partial least squares regression (PLSR), support vector regression (SVR), extreme learning machine (ELM), and convolutional neural network regression (CNNR), among others. However, for CNNR, only the raw spectra were used as the input. Out of the various models in this experiment, the SPA-CNNR model exhibited the most promising performances (RP = 0.839, RMSEP = 0.261, and RPD = 1.849).
Additionally, PLS-DA models for maturity discrimination of Camellia oleifera fruits were created using full wavelengths, characteristic wavelengths, and their fusion CMI, respectively. Of these models, the PLS-DA model using the fused data set achieved the highest maturity classification accuracy, with the best simplified model achieving 88.6% accuracy in a prediction set.
With these findings, it has been deemed possible to use a portable hyperspectral imager for in-field determination of the internal quality and maturity stages of Camellia oleifera fruits. Furthermore, this supports efforts for non-destructive quality inspection and timely harvesting of Camellia oleifera fruits.
(1) Yuan, W.; Zhou, H.; Zhou, Y.; Zhang, C.; Jiang, X.; Jiang, H. In-Field and Non-Destructive Determination of Comprehensive Maturity Index and Maturity Stages of Camellia oleifera Fruits Using a Portable Hyperspectral Imager. Spectrochim. Acta Part A: Mol. Biomol. Spectrosc. 2024, 315, 124266. DOI: 10.1016/j.saa.2024.124266
(2) Lavery, P. Hyperspectral Images of Fuji Apples Used as Predictive Data for Fruit Bruise Area. MJH Life Sciences 2024. https://www.spectroscopyonline.com/view/hyperspectral-images-fuji-apples-predictive-data-fruit-bruise-area (accessed 2024-4-18)
New Magnetic Flow Device Speeds Up Detection of Lactic Acid Bacteria and Yeast in Fermentation
November 11th 2024Researchers at Henan Agricultural University have developed a multi-channel magnetic flow device combined with surface-enhanced Raman spectroscopy (SERS) for the rapid and precise isolation, identification, and quantification of lactic acid bacteria and yeast, revolutionizing quality control in fermented food production.
Breakthrough Raman Spectroscopy Study Reveals Key to Advanced PFAS Detection
November 6th 2024Researchers from Sichuan University and the University of Georgia have developed an advanced method combining Raman spectroscopy and chemometric analysis to effectively identify and distinguish between various PFAS compounds, improving detection and environmental monitoring capabilities.