Integration of independently preprocessed spectra to improve classification of repeatedly fried durian chips using portable near-infrared spectrometer and machine learning

Arthit Phuangsombut; Kaewkarn Phuangsombut; Sirinad Noypitak; Anupun Terdwongworakul; Arthit Phuangsombut; Kaewkarn Phuangsombut; Sirinad Noypitak; Anupun Terdwongworakul

doi:10.3934/agrfood.2025029

AIMS Agriculture and Food

2025, Volume 10, Issue 3: 577-595. doi: 10.3934/agrfood.2025029

Previous Article Next Article

Research article

Integration of independently preprocessed spectra to improve classification of repeatedly fried durian chips using portable near-infrared spectrometer and machine learning

Department of Agricultural Engineering, Faculty of Engineering at Kamphaeng Saen, Kasetsart University, Kamphaeng Saen Campus, Nakhon Pathom, 73140, Thailand

Received: 16 March 2025 Revised: 12 June 2025 Accepted: 23 June 2025 Published: 31 July 2025

Durian chips fried in repeated-use oil can be harmful to health, as prolonged use of the oil leads to excessively high levels of total polar compounds. This study explored the use of a portable near-infrared (NIR) spectrometer combined with machine learning techniques to classify chips into two groups based on frying duration. An alternative fusion technique was proposed by integrating various independently preprocessed spectral data before model analysis. Binary classification was performed to distinguish between chips fried in oil used for 48 hours or less and those fried in oil used for more than that duration. Based on the results, the model developed using partial least squares discriminant analysis with fusion-preprocessed spectral data achieved superior classification performance, with a correlation coefficient of prediction (r_p) of 0.810 and a root mean square error of prediction (RMSEP) of 0.294, compared to the model using traditional preprocessed spectral data (r_p of 0.791 and RMSEP of 0.310). Subsequently, the fusion-preprocessed spectral data were used to develop a support vector machine (SVM) model, which demonstrated strong performance, achieving a precision of 91.9% in minimizing false positives and reducing the misclassification of chips fried in oil for a longer duration compared to those fried for a shorter duration. However, the SVM model lacked robustness when applied to different datasets. When trained and tested on the same dataset, the classification performance was highest for distinguishing between chips fried in oil used for 36 h or less and those fried in oil used for more than 36 h, achieving a precision of 97.8% on the test set. These findings suggested that the combination of a portable NIR spectrometer and the fusion preprocessing technique held promise for rapid, on-site assessment of chips fried in oil used for up to 36 h. Nonetheless, further studies should be undertaken to enhance the robustness and generalizability of the SVM model.
- durian chip,
- fried oil quality,
- portable device,
- near-infrared spectroscopy,
- fusion
Citation: Arthit Phuangsombut, Kaewkarn Phuangsombut, Sirinad Noypitak, Anupun Terdwongworakul. Integration of independently preprocessed spectra to improve classification of repeatedly fried durian chips using portable near-infrared spectrometer and machine learning[J]. AIMS Agriculture and Food, 2025, 10(3): 577-595. doi: 10.3934/agrfood.2025029

Related Papers:

Abstract

Durian chips fried in repeated-use oil can be harmful to health, as prolonged use of the oil leads to excessively high levels of total polar compounds. This study explored the use of a portable near-infrared (NIR) spectrometer combined with machine learning techniques to classify chips into two groups based on frying duration. An alternative fusion technique was proposed by integrating various independently preprocessed spectral data before model analysis. Binary classification was performed to distinguish between chips fried in oil used for 48 hours or less and those fried in oil used for more than that duration. Based on the results, the model developed using partial least squares discriminant analysis with fusion-preprocessed spectral data achieved superior classification performance, with a correlation coefficient of prediction (r_p) of 0.810 and a root mean square error of prediction (RMSEP) of 0.294, compared to the model using traditional preprocessed spectral data (r_p of 0.791 and RMSEP of 0.310). Subsequently, the fusion-preprocessed spectral data were used to develop a support vector machine (SVM) model, which demonstrated strong performance, achieving a precision of 91.9% in minimizing false positives and reducing the misclassification of chips fried in oil for a longer duration compared to those fried for a shorter duration. However, the SVM model lacked robustness when applied to different datasets. When trained and tested on the same dataset, the classification performance was highest for distinguishing between chips fried in oil used for 36 h or less and those fried in oil used for more than 36 h, achieving a precision of 97.8% on the test set. These findings suggested that the combination of a portable NIR spectrometer and the fusion preprocessing technique held promise for rapid, on-site assessment of chips fried in oil used for up to 36 h. Nonetheless, further studies should be undertaken to enhance the robustness and generalizability of the SVM model.

References

[1]	Sun Y, Zhang M, Fan D (2019) Effect of ultrasonic on deterioration of oil in microwave vacuum frying and prediction of frying oil quality based on low field nuclear magnetic resonance (LF-NMR). Ultrason Sonochem 51: 77–89. https://doi.org/10.1016/j.ultsonch.2018.10.015 doi: 10.1016/j.ultsonch.2018.10.015
[2]	Gertz C, Behmer D (2014) Application of FT‐NIR spectroscopy in assessment of used frying fats and oils. Eur J Lipid Sci Technol 116: 756–762. https://doi.org/10.1002/ejlt.201300270 doi: 10.1002/ejlt.201300270
[3]	Hassanien MFR, Sharoba AM (2014) Rheological characteristics of vegetable oils as affected by deep frying of French fries. Food Measure 8: 171–179. https://doi.org/10.1007/s11694-014-9178-3 doi: 10.1007/s11694-014-9178-3
[4]	Feng H, Li Y, Sui X, et al. (2016) Effect of frying cycles on polar components in soybean oil and absorbed lipids of fried potatoes. Trans Chin Soc Agric Eng 32: 309–314. https://doi.org/10.11975/j.issn.1002-6819.2016.03.044 doi: 10.11975/j.issn.1002-6819.2016.03.044
[5]	Sanchez-Muniz FJ, Cuesta C, Garrido-Polonio C (1993) Sunflower oil used for frying: combination of column, gas and high-performance size-exclusion chromatography for its evaluation. JAOCS 70: 235–240. https://doi.org/10.1007/BF02545301 doi: 10.1007/BF02545301
[6]	Hosseini H, Ghorbani M, Meshginfar N, et al. (2016) A review on frying: procedure, fat, deterioration progress and health hazards. JAOCS 93: 445–466. https://doi.org/10.1007/s11746-016-2791-z doi: 10.1007/s11746-016-2791-z
[7]	Ng CL, Wehling RL, Cuppett SL (2007) Method for determining frying oil degradation by near-infrared spectroscopy. J Agri Food Chem 55: 593–597. https://doi.org/10.1021/jf061841d doi: 10.1021/jf061841d
[8]	Ng CL, Wehling RL, Cuppett SL (2011) Near-infrared spectroscopic determination of degradation in vegetable oils used to fry various foods. J Agri Food Chem 59: 12286–12290. https://doi.org/10.1021/jf202740e doi: 10.1021/jf202740e
[9]	Öğütcü M, Aydeniz B, Büyükcan MB (2012) Determining frying oil degradation by near infrared spectroscopy using chemometric techniques. JAOCS 89: 1823–1830. https://doi.org/10.1007/s11746-012-2087-x doi: 10.1007/s11746-012-2087-x
[10]	Kazemi S, Wang N, Ngadi M, et al. (2005) Evaluation of frying oil quality using VIS/NIR hyperspectral analysis. Agric Eng Int: CIGR J Ⅶ: EP05 001.
[11]	Liu Y, Sun L, Li Y, et al. (2018) Quality evaluation of fried soybean oil base on near infrared spectroscopy. J Food Process Eng 41: e12887. https://doi.org/10.1111/jfpe.12887 doi: 10.1111/jfpe.12887
[12]	Liu Y, Sun L, Bai H, et al. (2020) Detection for frying times of various edible oils based on near-infrared spectroscopy. Appl Sci 10: 7789. https://doi.org/10.3390/app10217789 doi: 10.3390/app10217789
[13]	Liu Y, Sun L, Du C, et al. (2020) Near-infrared prediction of edible oil frying times based on Bayesian Ridge Regression. Optik 218: 164950. https://doi.org/10.1016/j.ijleo.2020.164950 doi: 10.1016/j.ijleo.2020.164950
[14]	Ma J, Zhang H, Tuchiya T, et al. (2014) Rapid determination of degradation of frying oil using near-infrared spectroscopy. Food Sci Technol Res 20: 217–223. https://doi.org/10.3136/fstr.20.217 doi: 10.3136/fstr.20.217
[15]	García-Martín JF, Alés-Álvarez FJ, del Carmen López-Barrera M, et al. (2019) Cetane number prediction of waste cooking oil-derived biodiesel prior to transesterification reaction using near infrared spectroscopy. Fuel 240: 10–15. https://doi.org/10.1016/j.fuel.2018.11.142 doi: 10.1016/j.fuel.2018.11.142
[16]	Isdhiyanti MIT, Budiastra IW, Mala DM, et al. (2024) Prediction of chemical contents of cooking oil using near-infrared spectroscopy. Food Res 8: 147–152. https://doi.org/10.26656/fr.2017.8(6).520 doi: 10.26656/fr.2017.8(6).520
[17]	Maniwara P, Meesombat R, Malang S, et al. (2024) Determination of oil quality during crispy pork rind frying: Near infrared spectra and color values as predictors. J Food Eng 383: 112251. https://doi.org/10.1016/j.jfoodeng.2024.112251 doi: 10.1016/j.jfoodeng.2024.112251
[18]	Palarea-Albaladejo J, Cayuela-Sánchez JA, Moriana-Correro E (2022) Estimating fat components of potato chips using visible and near-infrared spectroscopy and a compositional calibration model. Food Anal Methods 15: 133–143. https://doi.org/10.1007/s12161-021-02106-0 doi: 10.1007/s12161-021-02106-0
[19]	Pedreschi F, Segtnan VH, Knutsen SH (2010) On-line monitoring of fat, dry matter and acrylamide contents in potato chips using near infrared interactance and visual reflectance imaging. Food Chem 121: 616–620. https://doi.org/10.1016/j.foodchem.2009.12.075 doi: 10.1016/j.foodchem.2009.12.075
[20]	Areekij S, Ritthiruangdej P, Kasemsumran S, et al. (2017) Rapid and nondestructive analysis of deep-fried taro chip qualities using near infrared spectroscopy. J Near Infrared Spectrosc 25: 127–137. https://doi.org/10.1177/0967033516686655 doi: 10.1177/0967033516686655
[21]	López Maestresalas A, López Molina C, Oliva Lobo GA, et al. (2022) Evaluation of near-infrared hyperspectral imaging for the assessment of potato processing aptitude. Front Nutr 9: 999877. https://doi.org/10.3389/fnut.2022.999877 doi: 10.3389/fnut.2022.999877
[22]	Zhang Y, Wang Y (2023) Machine learning applications for multi-source data of edible crops: A review of current trends and future prospects. Food Chem: X, 100860. https://doi.org/10.1016/j.fochx.2023.100860 doi: 10.1016/j.fochx.2023.100860
[23]	Di Rosa AR, Leone F, Cheli F, Chiofalo V (2017) Fusion of electronic nose, electronic tongue and computer vision for animal source food authentication and quality assessment–A review. J Food Eng 210: 62–75. https://doi.org/10.1016/j.jfoodeng.2017.04.024 doi: 10.1016/j.jfoodeng.2017.04.024
[24]	Xia H, Chen W, Hu D, et al. (2024) Rapid discrimination of quality grade of black tea based on near-infrared spectroscopy (NIRS), electronic nose (E-nose) and data fusion. Food Chem 440: 138242. https://doi.org/10.1016/j.foodchem.2023.138242 doi: 10.1016/j.foodchem.2023.138242
[25]	Hong Y, Birse N, Quinn B, et al. (2023) Data fusion and multivariate analysis for food authenticity analysis. Nat Commun 14: 3309. https://doi.org/10.1038/s41467-023-38382-z doi: 10.1038/s41467-023-38382-z
[26]	Xu P, Fu L, Xu K, et al. (2023) Investigation into maize seed disease identification based on deep learning and multi-source spectral information fusion techniques. J Food Compos Anal 119: 105254. https://doi.org/10.1016/j.jfca.2023.105254 doi: 10.1016/j.jfca.2023.105254
[27]	An H, Zhai C, Zhang F, et al. (2023) Quantitative analysis of Chinese steamed bread staling using NIR, MIR, and Raman spectral data fusion. Food Chem 405:134821.
[28]	Cevoli C, Iaccheri E, Fabbri A, et al. (2024) Data fusion of FT-NIR spectroscopy and Vis/NIR hyperspectral imaging to predict quality parameters of yellow flesh "Jintao" kiwifruit. Biosyst Eng 237: 157–169. https://doi.org/10.1016/j.biosystemseng.2023.12.011 doi: 10.1016/j.biosystemseng.2023.12.011
[29]	Ping J, Ying Z, Hao N, et al. (2024) Rapid and non-destructive identification of Panax ginseng origins using hyperspectral imaging, visible light imaging, and X-ray imaging combined with multi-source data fusion strategies. Int Food Res 192: 114758. https://doi.org/10.1016/j.foodres.2024.114758 doi: 10.1016/j.foodres.2024.114758
[30]	Li X, Cai M, Li M, et al. (2023) Combining Vis-NIR and NIR hyperspectral imaging techniques with a data fusion strategy for the rapid qualitative evaluation of multiple qualities in chicken. Food Control 145: 109416. https://doi.org/10.1016/j.foodcont.2022.109416 doi: 10.1016/j.foodcont.2022.109416
[31]	Feng Y, Dong F, Chen Y, et al. (2024) Combining Vis-NIR and NIR hyperspectral imaging techniques with a data fusion strategy for prediction of norfloxacin residues in mutton. Spectrochim Acta A Mol Biomol Spectrosc 322: 124844. https://doi.org/10.1016/j.saa.2024.124844 doi: 10.1016/j.saa.2024.124844
[32]	Posom J, Phuphaphud A, Saengprachatanarug K, et al. (2022) Real-time measuring energy characteristics of cane bagasse using NIR spectroscopy. Sens BioSensing Res 38: 100519. https://doi.org/10.1016/j.sbsr.2022.100519 doi: 10.1016/j.sbsr.2022.100519
[33]	Leardi R (2000) Application of genetic algorithm–PLS for feature selection in spectral data sets. J Chemom 14: 643–655. https://doi.org/10.1002/1099-128X(200009/12)14:5/6%3C643::AID-CEM621%3E3.0.CO;2-E doi: 10.1002/1099-128X(200009/12)14:5/6%3C643::AID-CEM621%3E3.0.CO;2-E
[34]	Haupt, SE (2009) Introduction to genetic algorithms, In: Artificial Intelligence Methods in the Environmental Sciences, 103–125, Dordrecht: Springer Netherlands. https://doi.org/10.1007/978-1-4020-9119-3_5
[35]	Ali MM, Hashim, N, Shahamshah, MI (2021) Durian (Durio zibethinus) ripeness detection using thermal imaging with multivariate analysis. Postharvest Biol Technol 176: 111517. https://doi.org/10.1016/j.postharvbio.2021.111517 doi: 10.1016/j.postharvbio.2021.111517
[36]	Ditcharoen S, Sirisomboon P, Saengprachatanarug K, et al. (2023) Improving the non-destructive maturity classification model for durian fruit using near-infrared spectroscopy. Artif Intell Agric 7: 35–43. https://doi.org/10.1016/j.aiia.2023.02.002 doi: 10.1016/j.aiia.2023.02.002
[37]	Smola AJ, Schölkopf B (2004) A tutorial on support vector regression. Stat Comput 14: 199–222. https://doi.org/10.1023/B:ASTCO.0000035301.49549.88 doi: 10.1023/B:ASTCO.0000035301.49549.88
[38]	Raschka S, Mirjalili V (2019) Python machine learning: Machine learning and deep learning with Python, scikit-learn, and TensorFlow 2nd Edition, Birmingham: Packt Publishing.
[39]	Osborne BG, Fearn T, Hindle PH (1993) Practical NIR Spectroscopy with Application in Food and BeVerage Analysis, 2 Eds., Essex: Longman Scientific & Technical, 13–35.
[40]	Workman J, Weyer L (2012) Practical Guide and Spectral Atlas for Interpretive Near Infrared Spectroscopy, 2 Eds., Boca Raton: CRC Press. https://doi.org/10.1201/b11894

Reader Comments

Your name:*

Email:*
© 2025 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)