KMITL

Permanent URI for this communityhttps://dspace.kmitl.ac.th/handle/123456789/1

Browse

Search Results

Now showing 1 - 10 of 27

Some of the metrics are blocked by your
consent settings
Systematic evaluation of one-dimensional-to-two-dimensional near-infrared spectroscopy transformations with deep learning for quantifying coconut sap adulteration
(2026-10-01)
Lapcharoensuk, Ravipat
;
Sitorus, Agustami
Near-infrared (NIR) spectroscopy have limitations when combined with deep learning (DL) algorithms because they rely on low-dimensional datasets. Therefore, we investigated the potential of transforming one-dimensional (1D) NIR spectra into two-dimensional (2D) spectrograms using synchronous and asynchronous techniques and the continuous wavelet transform (CWT) and their effectiveness by integrating with DL for detecting adulteration in coconut sap. NIR spectra (12,500–4000 cm−1) were collected from binary mixtures (0%–100%; w /w). The performance of all DL (convolutional neural networks-CNN, AlexNet and ResNet) models was compared with that of partial least squares (PLS). The models were ranked in the mentioned order based on their performances: 2D-CWT ' 2D-asynchronous ' 2D-synchronous ' 1D/2D-PLS. The important features of the best model can be explained and visualized using gradient-weighted-class-activation-mapping. The findings highlight that the 1D-to-2D NIR data transformation combined with DL is a highly robust approach because it addresses the feature representation gap in NIR data and effectively captures the spatial–spectral correlations.
Some of the metrics are blocked by your
consent settings
Low-cost Multispectral Acquisition Device Coupled with Machine Learning for Detecting Adulteration of Honey
(2026-07-15)
Boodnon, Wutthiphong
;
Lunvongsa, Thayanont
;
Suntisakoonwong, Phanchay
;
Sitorus, Agustami
;
Lapcharoensuk, Ravipat
Honey is a natural sweetener created by honeybees from the nectar of flowers. Honey's extensive health benefits have led to its widespread use across multiple industries. Honey adulteration with inferior substances undermines its quality, reducing natural nutrients and antioxidants, and diminishing its health benefits. This study aimed to study the possibility of detection of honey adulteration with a low-cost multispectral device coupled with machine learning. The adulterated honey came from deliberate adulteration with cane syrup in the 1 to 90% range. Spectral data was collected for pure honey and the adulterated honey samples at the wavelengths of 610, 680, 730, 760, 810, and 860 nm. The detection models for distinguishing pure and adulterated honey were developed by Linear Discriminant Analysis (LDA), Partial Least Squares Discriminant Analysis (PLS-DA), C-Support Vector Machine (C-SVM), and K-Nearest Neighbors (KNN). All models achieved high accuracy between 0.91 and 0.98 and maintained balanced precision and recall metrics. This study serves as a guideline for developing a low-cost portable honey authentication device that is practical for real-world applications.
Some of the metrics are blocked by your
consent settings
Interpretable Machine Learning for Sugarcane Harvester Performance: A Comparison of Additive and Tree-Based Models on Telematics Data
(2026-07-01)
Kaewkabthong, Apidul
;
Saijai, Jedsada
;
Sriphuk, Pisitwitthaya
;
Sitorus, Agustami
;
Udompetaikul, Vasu
Sugarcane harvester performance varies substantially with field geometry, crop, and operator factors, yet separating these sources from telematics data while preserving engineering interpretability remains a methodological gap. This study models field efficiency (Eff) and harvesting capacity (C<inf>a</inf>) separately from JDLink telematics, aligning model structure with each target’s response behavior. Operational data covered 105 plots across four seasons (2019/20–2022/23) from three John Deere CH570 chopper harvesters in eastern Thailand. Six engineering-relevant predictors were retained after multicollinearity screening, and linear (MLR), additive nonlinear (GAM), and tree-based models were compared under 5-fold grouped cross-validation by BaseField (87 groups). Eff was assigned to GAM (R2<inf>CV</inf> = 0.621 ± 0.114) on the basis of its threshold-like response to turning frequency; C<inf>a</inf> was retained for MLR (R2<inf>CV</inf> = 0.681 ± 0.121), with GAM essentially tied. Train–validation gaps were substantially smaller for additive models (0.096–0.118) than for tuned tree-based candidates (GBR 0.210–0.302, RF 0.322–0.358). Turning frequency (TF) and perimeter-to-area ratio (PAR) were the strongest predictors, and a constant-turn-time partial-out test indicated that TF’s univariate effect on Eff is largely mediated by the time-budget identity. Tactical interventions (path planning, operator training, machine–field allocation) are immediately feasible, although strategic field-layout change remains constrained by smallholder land tenure.
Some of the metrics are blocked by your
consent settings
Quantitative analysis based on image processing combined with machine learning and deep learning to determine the adulteration in nutmeg powder
(2025-10-01)
Sitorus, Agustami
;
Pambudi, Suluh
;
Boodnon, Wutthiphong
;
Lapcharoensuk, Ravipat
The global demand for nutmeg powder has increased, raising the risk of adulteration and necessitating an efficient and cost-effective screening method. The objective of this study is to develop a calibration model to predict the adulteration of nutmeg powder by cinnamon powder using a novel approach by integrating image processing with machine learning (ML) and deep learning (DL). Eight regressors, including four ML regressors (multiple linear ridge regression–MLRR, partial least squares regression–PLSR, multi-layer perceptron–MLP, and adaptive boosting–ABS) and four DL regressors (convolutional neural networks–CNN, AlexNET, residual networks–ResNET, and GoogleNET), were employed to analyze 1800 images of adulteration samples ranging from 0 % to 35 % (w/w). Among ML models, MLP achieved the highest accuracy in prediction (R<inf>p</inf>²=0.922, RMSEP=2.804 %, RPD=3.59), while PLSR (R<inf>p</inf>²=0.876, RMSEP=3.538 %, RPD=2.84), MLRR (R<inf>p</inf>²=0.872, RMSEP=3.596 %, RPD=2.80), and ABS (R<inf>p</inf>²=0.849, RMSEP=3.904 %, RPD=2.58) underperformed. For DL, ResNET (R<inf>p</inf>²=0.882, RMSEP=3.416 %, RPD=2.91) surpassed CNN (R<inf>p</inf>²=0.876, RMSEP=3.505 %, RPD=2.84), AlexNET (R<inf>p</inf>²=0.801, RMSEP=4.429 %, RPD=2.24), and GoogleNET (R<inf>p</inf>²=0.751, RMSEP=4.963 %, RPD=2.00). The MLP's superiority highlights its compatibility with ORB-based feature extraction for nonlinear adulteration patterns, outperforming complex DL architectures. This research highlights the potential of image processing supported by ML and DL as a rapid and low-cost tool for future nutmeg powder adulteration screening.
Some of the metrics are blocked by your
consent settings
Development of automatic tuning for combined preprocessing and hyperparameters of machine learning and its application to NIR spectral data of coconut milk adulteration
(2024-11-01)
Sitorus, Agustami
;
Lapcharoensuk, Ravipat
This study proposed a novel approach to automatically select the preprocessing methods and hyperparameters of machine learning (ML) algorithms based on their best performance in cross-validation for near-infrared (NIR) spectroscopy data. The proposed method simultaneously incorporates single or multiple-preprocessing steps and tunes hyperparameters to determine the best model performance for FT-NIR and Micro-NIR spectral data of coconut milk adulteration with distilled water and mature coconut water in the range of 0%–50%. Computational experiments were conducted using nine single preprocessing types, three types of ML classifier (linear discriminant analysis (LDA), k-nearest neighbour (KNN), multilayer perceptron (MLP)) and three types of ML regressor (partial least squares (PLS), KNN, MLP). The proposed performance strategy effectively addressed and produced satisfactory outcomes for classification and regression challenges in coconut milk adulteration. Finally, the results demonstrated that the proposed approach can more accurately determine the best model, particularly for NIR spectroscopy of coconut milk adulteration.
Some of the metrics are blocked by your
consent settings
Discrimination model of geographical area from coconut milk by near-infrared spectroscopy: Exploration in tandem with classical chemometrics, machine learning, and deep learning
(2024-11-01)
Sitorus, Agustami
;
Lapcharoensuk, Ravipat
This work proposes exploring the discrimination model by near-infrared (NIR) spectroscopy (FT-NIR and Micro-NIR) for geographical source areas of coconut milk in tandem with the classical to modern chemometrics classifier. The discrimination model was developed using qualitative chemometrics techniques from classic (Principal Component Analysis-PCA, Partial Least Squares Discriminant Analysis-PLS-DA, Linear Discriminant Analysis-LDA) to modern, including classifiers from machine learning (Support Vector Machine-SVM, k-Nearest Neighbor-KNN, Artificial Neural Network-ANN) and deep learning (Simple Convolutional Neural Networks-S-CNN, S-AlexNET, Residual Networks-ResNET). Three sources as geographical areas of coconut milk originally from Thailand were used, including the south region (Chumphon Province), middle region (Samut Songkhram Province), and east region (Chonburi Province). Our findings showed that a classifier from SVM and ResNET could yield the optimal performance for discriminating the geographical source area of coconut milk using FT-NIR. Furthermore, when using Micro-NIR, the classifier from LDA, SVM, KNN and ResNET delivered the highest accuracy. The performance discrimination models above were excellent when classified based on the kappa coefficient. This study concluded that both FT-NIR and Micro-NIR supported by classical to modern chemometric classifiers could be used to evaluate the geographical area source from coconut milk. Also, the method in this study includes a strategy for discovering feature-important NIR spectra for interpretability purposes, thereby facilitating the qualitative interpretation of results for all types of classifiers.
Some of the metrics are blocked by your
consent settings
Rapid Classification of Sugarcane Nodes and Internodes Using Near-Infrared Spectroscopy and Machine Learning Techniques
(2024-11-01)
Veerasakulwat, Siramet
;
Sitorus, Agustami
;
Udompetaikul, Vasu
Accurate and rapid discrimination between nodes and internodes in sugarcane is vital for automating planting processes, particularly for minimizing bud damage and optimizing planting material quality. This study investigates the potential of visible-shortwave near-infrared (Vis–SWNIR) spectroscopy (400–1000 nm) combined with machine learning for this classification task. Spectral data were acquired from the sugarcane cultivar Khon Kaen 3 at multiple orientations, and various preprocessing techniques were employed to enhance spectral features. Three machine learning algorithms, linear discriminant analysis (LDA), K-Nearest Neighbors (KNNs), and artificial neural networks (ANNs), were evaluated for their classification performance. The results demonstrated high accuracy across all models, with ANN coupled with derivative preprocessing achieving an F1-score of 0.93 on both calibration and validation datasets, and 0.92 on an independent test set. This study underscores the feasibility of Vis–SWNIR spectroscopy and machine learning for rapid and precise node/internode classification, paving the way for automation in sugarcane billet preparation and other precision agriculture applications.
Some of the metrics are blocked by your
consent settings
Exploring Deep Learning to Predict Coconut Milk Adulteration Using FT-NIR and Micro-NIR Spectroscopy
(2024-04-01)
Sitorus, Agustami
;
Lapcharoensuk, Ravipat
Accurately identifying adulterants in agriculture and food products is associated with preventing food safety and commercial fraud activities. However, a rapid, accurate, and robust prediction model for adulteration detection is hard to achieve in practice. Therefore, this study aimed to explore deep-learning algorithms as an approach to accurately identify the level of adulterated coconut milk using two types of NIR spectrophotometer, including benchtop FT-NIR and portable Micro-NIR. Coconut milk adulteration samples came from deliberate adulteration with corn flour and tapioca starch in the 1 to 50% range. A total of four types of deep-learning algorithm architecture that were self-modified to a one-dimensional framework were developed and tested to the NIR dataset, including simple CNN, S-AlexNET, ResNET, and GoogleNET. The results confirmed the feasibility of deep-learning algorithms for predicting the degree of coconut milk adulteration by corn flour and tapioca starch using NIR spectra with reliable performance (R2 of 0.886–0.999, RMSE of 0.370–6.108%, and Bias of −0.176–1.481). Furthermore, the ratio of percent deviation (RPD) of all algorithms with all types of NIR spectrophotometers indicates an excellent capability for quantitative predictions for any application (RPD > 8.1) except for case predicting tapioca starch, using FT-NIR by ResNET (RPD < 3.0). This study demonstrated the feasibility of using deep-learning algorithms and NIR spectral data as a rapid, accurate, robust, and non-destructive way to evaluate coconut milk adulterants. Last but not least, Micro-NIR is more promising than FT-NIR in predicting coconut milk adulteration from solid adulterants, and it is portable for in situ measurements in the future.
Some of the metrics are blocked by your
consent settings
Near-Infrared Spectroscopy with Machine Learning for Classifying and Quantifying Nutmeg Adulteration
(2024-01-01)
Sitorus, Agustami
;
Pambudi, Suluh
;
Boodnon, Wutthiphong
;
Lapcharoensuk, Ravipat
Near-infrared spectroscopy (NIRS) provides broadbands, overtones, and combinations of organic-bond vibrations and has been used to characterize agricultural and food products. The adulteration of grated nutmeg with cinnamon is extremely profitable and difficult to detect; to prevent retail fraud, it is vital to differentiate between these materials. This study proposes a model for classifying the adulteration of nutmeg with cinnamon and predicting the level of adulteration. NIR spectra were characterized with six machine learning (ML) algorithms, namely, the principal component-multilayer perceptron (PC-MLP), principal component-linear discriminant analysis (PC-LDA), partial least squares regression (PLSR), support vector machine (SVM), random forest (RF), and decision tree (DT) methods. PC-MLP provided 100% accuracy in calibration and prediction in distinguishing nutmeg from cinnamon. In addition, this approach showed excellent performance in predicting the adulteration ratio of nutmeg and cinnamon with a high coefficient of determination of prediction (R 2<inf>pred</inf>) value of 0.9969, low root mean square error of prediction (RMSEP) value of 0.5728%, and high ratio of prediction to deviation (RPD) value of 17.9605. Therefore, this study indicates the potential of integrating NIR spectroscopy with PC-MLP to classify and quantify the adulteration of nutmeg.
Some of the metrics are blocked by your
consent settings
Classification of the Crosslink Density Level of Para Rubber Thick Film of Medical Glove by Using Near-Infrared Spectral Data
(2024-01-01)
Jongyingcharoen, Jiraporn Sripinyowanich
;
Howimanporn, Suppakit
;
Sitorus, Agustami
;
Phanomsophon, Thitima
;
Posom, Jetsada
Classification of the crosslink density level of para rubber medical gloves by using near-infrared spectral data combined with machine learning is the first time reported in this paper. The spectra of medical glove samples with different crosslink densities acquired by an ultra-compact portable MicroNIR spectrometer were correlated with their crosslink density levels, which were referencely evaluated by the toluene swell index (TSI). The machine learning protocols used to classify the 3 groups of TSI were specified as less than 80% TSI, 80–88% TSI, and more than 88% TSI. The 80–88% TSI group was the group in which the compounded latex was suitable for medical glove production, which made the glove specification comply with the requirements of customers as indicated by the tensile test. The results show that when comparing the algorithms used for modeling, the linear discriminant analysis (LDA) developed by 2nd derivative spectra with 15 k-best selected wavelengths fairly accurately predicted the class but was most reliable among other algorithms, i.e., artificial neural networks (ANN), support vector machines (SVM), and k-nearest neighbors (kNN), due to higher prediction accuracy, precision, recall, and F1-score of the same value of 0.76 and no overfitting or underfitting prediction. This developed model can be implemented in the glove factory for screening purposes in the production line. However, deep learning modeling should be explored with a larger sample number required for better model performance.

KMITL

Browse

Filters

Settings

Sort By

Results per page

Search Results