Non-Contact Multispectral Image Binary Classification of Water and Sodium Hydroxide Solutions Using Convolutional Neural Networks

Siti Rusdiana; Asep  Rusyana; Juwita; Mauliza Putri; Aufa Rafiki; Souvik Das

doi:10.35882/jeeemi.v8i3.1727

Siti Rusdiana Department of Mathematics, Faculty of Mathematics and Natural Sciences, Universitas Syiah Kuala, Banda Aceh, Aceh, Indonesia https://orcid.org/0000-0002-3501-8521
Asep Rusyana Department of Statistics, Faculty of Mathematics and Natural Sciences, Universitas Syiah Kuala, Banda Aceh, 23111, Aceh, Indonesia. https://orcid.org/0000-0002-4399-0281
Juwita Department of Informatics, Faculty of Mathematics and Natural Sciences, Universitas Syiah Kuala, Banda Aceh, 23111, Aceh, Indonesia. https://orcid.org/0009-0009-7179-6476
Mauliza Putri Department of Electrical and Computer Engineering, Faculty of Engineering, Universitas Syiah Kuala, Banda Aceh, 23111, Aceh, Indonesia. https://orcid.org/0009-0009-5634-9555
Aufa Rafiki Department of Electrical and Computer Engineering, Faculty of Engineering, Universitas Syiah Kuala, Banda Aceh, 23111, Aceh, Indonesia. https://orcid.org/0009-0008-3031-8008
Souvik Das Department of Industrial Design, National Institute of Technology Rourkela, Rourkela 769008, India. https://orcid.org/0000-0001-8296-3289

DOI: https://doi.org/10.35882/jeeemi.v8i3.1727

Keywords: Multispectral Imaging; Transparent Liquid Classification; Sodium Hydroxide Solution; Convolutional Neural Networks; Optimization.

Abstract

Identification of visually transparent liquids remains a challenging problem in non-contact sensing because chemically different solutions can appear nearly identical under normal observation. In laboratory and industrial environments, direct-contact chemical measurements are reliable but may require sample handling, probe calibration, cleaning, and additional processing time, which can limit their use in rapid or automated monitoring systems. This study aims to develop and evaluate a non-contact image-based classification framework for distinguishing pure water (H₂O) from sodium hydroxide solution (H₂O with NaOH) using multispectral fluctuation-pattern images. The proposed approach integrates image preprocessing, K-means segmentation, and a convolutional neural network (CNN)-based classification. A balanced dataset of 1,050 multispectral images, consisting of 525 images for each class, was used in the experiment. Each image was resized, converted to grayscale, normalized, and segmented using K-means clustering to emphasize the dominant liquid-region fluctuation pattern before classification. Three CNN architectures, namely InceptionV3, VGG19, and DenseNet201, were trained and compared under identical data-splitting and evaluation conditions. The experimental results showed that VGG19 achieved the best testing performance, with an accuracy of 97.47%, precision of 95.18%, recall of 100.00%, and F1-score of 97.53%. DenseNet201 obtained 94.30% accuracy, while InceptionV3 achieved 89.24% accuracy. These results indicate that multispectral fluctuation-pattern images contain discriminative optical information that can be learned effectively by CNN models, even when the liquid samples are visually indistinguishable to the human eye. The proposed framework demonstrates the feasibility of non-contact transparent liquid identification and may support the development of automated monitoring systems for laboratory, chemical, and industrial applications where direct sample contact is undesirable or impractical.

Downloads

Download data is not yet available.

Author Biographies

Asep Rusyana, Department of Statistics, Faculty of Mathematics and Natural Sciences, Universitas Syiah Kuala, Banda Aceh, 23111, Aceh, Indonesia.

Asep Rusyana graduated from the Department of Statistics, IPB University, in 2000 with a Bachelor of Science. Then he continued to the same department and graduated in 2005 with a Master of Science. Then the author continued studying at the Doctoral Program in Study Program of Statistics and Data Science from 2019 until 2024. He became a lecturer at the Department of Mathematics, Universitas Syiah Kuala (USK), Banda Aceh, from 2006 to 2012. Then he became a lecturer at the Department of Statistics, USK from 2013 until now. During his time as a USK lecturer, the author was active as head of the computational statistics laboratory, chair of the Mathematics Department from January 2012 to December 2015, and staff at the USK quality assurance agency between 2008 and 2018. The author is also a member of IEEE since January 2023. While teaching at USK, the author teaches several courses including machine learning, computational statistics, algorithm and programing, experimental design, matrix algebra, multivariate analysis, introductory statistics, operations research, time series analysis, and calculus. More than 50 publications in applied statistics and data sciences have been published in journals or conferences at national and international levels. The author's areas of research interest are data science and applied statistics.

Juwita, Department of Informatics, Faculty of Mathematics and Natural Sciences, Universitas Syiah Kuala, Banda Aceh, 23111, Aceh, Indonesia.

Juwita is a senior lecturer in the Department of Informatics at Universitas Syiah Kuala, Indonesia. She earned her Ph.D. in Computer Science and Software Engineering from the University of Western Australia in 2025. Her recent research has primarily focused on medical image segmentation. In addition, her research interests encompass broader multidisciplinary domains, particularly in the development of information systems aimed at enhancing human quality of life, including applications in education and healthcare. She can be contacted at juwita@usk.ac.id.

Mauliza Putri, Department of Electrical and Computer Engineering, Faculty of Engineering, Universitas Syiah Kuala, Banda Aceh, 23111, Aceh, Indonesia.

Mauliza Putri was born on December 27, 2002, in Pulau Tiga. She is a student majoring in Electrical and Computer Engineering at Syiah Kuala University. Her undergraduate studies focused on multimedia technology, and her research explored the Comparative Performance Analysis of H₂O and H₂O–NaOH Mixtures Based on K-Means Segmentation, CNN, and Decision Fusion Classification. She actively participates in class and continues to develop her knowledge in her field. Enrolled in the class of 2022, she is committed to expanding her expertise and gaining practical experience. Her academic journey reflects her dedication to both theoretical and applied learning, preparing her to contribute to technological advancement in her field. She can be reached at mauliza6@mhs.usk.ac.id.

Aufa Rafiki, Department of Electrical and Computer Engineering, Faculty of Engineering, Universitas Syiah Kuala, Banda Aceh, 23111, Aceh, Indonesia.

Aufa Rafiki was born on April 20, 2003, in Banda Aceh, Indonesia. He received his Bachelor’s degree in Computer Engineering from Universitas Syiah Kuala in 2025 and graduated as the best graduate of the Faculty of Engineering. He also completed a fast-track Master’s degree in Electrical Engineering with a biomedical engineering concentration at Universitas Syiah Kuala. His research interests include EEG signal processing, artificial intelligence, deep learning, biomedical signal processing, and early detection of autism spectrum disorder using medical technology. He has experience in collaborative research, scientific publication, AI-based system development, and intellectual property outputs. He has also served as a research assistant and laboratory assistant. He can be contacted at aufa35@mhs.usk.ac.id.

Souvik Das, Department of Industrial Design, National Institute of Technology Rourkela, Rourkela 769008, India.

Souvik Das completed his B.Tech. in Electrical Engineering from RCC Institute of Information Technology (RCCIIT), West Bengal, India in 2014. He then obtained his M.Tech. in Industrial Engineering and Management from the Indian Institute of Technology (IIT) Kharagpur in 2017. He earned his Ph.D. in Industrial and Systems Engineering from IIT Kharagpur in 2023, where his research focused on data-driven modeling of cognitive workload using eye-tracking metrics. During his doctoral studies, he worked as a Principal Research Scientist at the Centre of Excellence in Safety Engineering and Analytics (CoE-SEA), IIT Kharagpur. He later served as a Postdoctoral Research Fellow (Visiting Scientist) at Purdue University, USA, working on AI-based analysis of accident and incident narratives for safety analytics. Since August 2024, he has been serving as an Assistant Professor in the Department of Industrial Design at the National Institute of Technology (NIT) Rourkela, India. His research interests include human factors, safety analytics, eye-tracking technology, virtual reality, and artificial intelligence for safety and design applications. He can be contacted at dass@nitrkl.ac.in.

References

L. Lajoie, A. S. Fabiano-Tixier, and F. Chemat, “Water as Green Solvent: Methods of Solubilisation and Extraction of Natural Products—Past, Present and Future Solutions,” Pharmaceuticals, vol. 15, no. 12, Art. no. 1507, 2022, doi: 10.3390/ph15121507.

C. D. Rodríguez-Fernández, L. M. Varela, C. Schröder, and E. L. Lago, “Charge delocalization and hyperpolarizability in ionic liquids,” J. Mol. Liq., vol. 349, Art. no. 118153, 2022, doi: 10.1016/j.molliq.2021.118153.

Y. Wu, H. Ye, Y. Yang, Z. Wang, and S. Li, “Liquid Content Detection in Transparent Containers: A Benchmark,” Sensors, vol. 23, no. 15, Art. no. 6656, 2023, doi: 10.3390/s23156656.

I. Shaw and P. Magee, “Acid–base quantification: A review of developing technology,” BJA Educ., vol. 22, no. 11, pp. 440–447, 2022, doi: 10.1016/j.bjae.2022.07.006.

W. Zhang et al., “Non-contact measurement method of liquid composition using microwave radar cross-section,” Sci. Rep., vol. 14, no. 1, Art. no. 29744, 2024, doi: 10.1038/s41598-024-81043-4.

J.-H. Leung et al., “Water pollution classification and detection by hyperspectral imaging,” Opt. Express, vol. 32, no. 14, pp. 23956–23965, 2024, doi: 10.1364/OE.522932.

S. Kendler, Z. Mano, R. Aharoni, R. Raich, and B. Fishbain, “Hyperspectral imaging for chemicals identification: A human-inspired machine learning approach,” Sci. Rep., vol. 12, no. 1, Art. no. 17580, 2022, doi: 10.1038/s41598-022-22468-7.

J. Zhu, J. Bao, and Y. Tao, “A Nondestructive Methodology for Determining Chemical Composition of Salvia miltiorrhiza via Hyperspectral Imaging Analysis and Squeeze-and-Excitation Residual Networks,” Sensors, vol. 23, no. 23, Art. no. 9345, 2023, doi: 10.3390/s23239345.

L. C. O. Tiong et al., “Machine vision-based detections of transparent chemical vessels toward the safe automation of material synthesis,” npj Comput. Mater., vol. 10, no. 1, Art. no. 42, 2024, doi: 10.1038/s41524-024-01216-7.

X. Zhao, L. Wang, Y. Zhang, X. Han, M. Deveci, and M. Parmar, “A review of convolutional neural networks in computer vision,” Artif. Intell. Rev., vol. 57, no. 4, Art. no. 99, 2024, doi: 10.1007/s10462-024-10721-6.

Z. Khan and J. Yang, “Nonparametric K-means clustering-based adaptive unsupervised colour image segmentation,” Pattern Anal. Appl., vol. 27, no. 1, Art. no. 17, 2024, doi: 10.1007/s10044-024-01228-5.

M. Sabha and M. Saffarini, “Selecting optimal k for K-means in image segmentation using GLCM,” Multimed. Tools Appl., vol. 83, no. 18, pp. 55587–55603, 2024, doi: 10.1007/s11042-023-17615-9.

H. Mittal, A. C. Pandey, M. Saraswat, S. Kumar, R. Pal, and G. Modwel, “A comprehensive survey of image segmentation: Clustering methods, performance parameters, and benchmark datasets,” Multimed. Tools Appl., vol. 81, no. 24, pp. 35001–35026, 2022, doi: 10.1007/s11042-021-10594-9.

M. Krichen, “Convolutional Neural Networks: A Survey,” Computers, vol. 12, no. 8, Art. no. 151, 2023, doi: 10.3390/computers12080151.

M. M. Taye, “Theoretical Understanding of Convolutional Neural Network: Concepts, Architectures, Applications, Future Directions,” Computation, vol. 11, no. 3, Art. no. 52, 2023, doi: 10.3390/computation11030052.

J. Ran, G. Dong, F. Yi, L. Li, and Y. Wu, “Automatic Measurement of Comprehensive Skin Types Based on Image Processing and Deep Learning,” Electronics, vol. 14, no. 1, Art. no. 49, 2025, doi: 10.3390/electronics14010049.

I. A. Kandhro et al., “Performance evaluation of E-VGG19 model: Enhancing real-time skin cancer detection and classification,” Heliyon, vol. 10, no. 10, Art. no. e31488, 2024, doi: 10.1016/j.heliyon.2024.e31488.

M. Bundea and G. M. Danciu, “Pneumonia Image Classification Using DenseNet Architecture,” Information, vol. 15, no. 10, Art. no. 611, 2024, doi: 10.3390/info15100611.

B. Dey, J. Ferdous, R. Ahmed, and J. Hossain, “Assessing deep convolutional neural network models and their comparative performance for automated medicinal plant identification from leaf images,” Heliyon, vol. 10, no. 1, Art. no. e23655, 2024, doi: 10.1016/j.heliyon.2023.e23655.

H. Zhou, X. Wang, K. Xia, Y. Ma, and G. Yuan, “Transfer Learning-Based Hyperspectral Image Classification Using Residual Dense Connection Networks,” Sensors, vol. 24, no. 9, Art. no. 2664, 2024, doi: 10.3390/s24092664.

R. Hou, J. Y. Lo, J. R. Marks, E. S. Hwang, and L. J. Grimm, “Classification performance bias between training and test sets in a limited mammography dataset,” PLOS ONE, vol. 19, no. 2, Art. no. e0282402, 2024, doi: 10.1371/journal.pone.0282402.

Z. Yang, R. O. Sinnott, J. Bailey, and Q. Ke, “A survey of automated data augmentation algorithms for deep learning-based image classification tasks,” Knowl. Inf. Syst., vol. 65, no. 7, pp. 2805–2861, 2023, doi: 10.1007/s10115-023-01853-2.

S. Seoni et al., “All you need is data preparation: A systematic review of image harmonization techniques in multi-center/device studies for medical support systems,” Comput. Methods Programs Biomed., vol. 250, Art. no. 108200, 2024, doi: 10.1016/j.cmpb.2024.108200.

F. Hu et al., “Image harmonization: A review of statistical and deep learning methods for removing batch effects and evaluation metrics for effective harmonization,” NeuroImage, vol. 274, Art. no. 120125, 2023, doi: 10.1016/j.neuroimage.2023.120125.

A. M. Ikotun, A. E. Ezugwu, L. Abualigah, B. Abuhaija, and J. Heming, “K-means clustering algorithms: A comprehensive review, variants analysis, and advances in the era of big data,” Inf. Sci., vol. 622, pp. 178–210, 2023, doi: 10.1016/j.ins.2022.11.139.

X. Chai, Z. Wu, W. Li, H. Fan, X. Sun, and J. Xu, “Image Segmentation Based on the Optimized K-Means Algorithm with the Improved Hybrid Grey Wolf Optimization: Application in Ore Particle Size Detection,” Sensors, vol. 25, no. 9, Art. no. 2785, 2025, doi: 10.3390/s25092785.

H. T. Lee, H. R. Cheon, S. H. Lee, M. Shim, and H. J. Hwang, “Risk of data leakage in estimating the diagnostic performance of a deep-learning-based computer-aided system for psychiatric disorders,” Sci. Rep., vol. 13, no. 1, Art. no. 16633, 2023, doi: 10.1038/s41598-023-43542-8.

O. Rainio, J. Teuho, and R. Klén, “Evaluation metrics and statistical tests for machine learning,” Sci. Rep., vol. 14, no. 1, Art. no. 6086, 2024, doi: 10.1038/s41598-024-56706-x.

L. Chen et al., “An Adaptive Parameter Optimization Deep Learning Model for Energetic Liquid Vision Recognition Based on Feedback Mechanism,” Sensors, vol. 24, no. 20, Art. no. 6733, 2024, doi: 10.3390/s24206733.

M. A. Lones, “Avoiding common machine learning pitfalls,” Patterns, vol. 5, no. 10, Art. no. 101046, 2024, doi: 10.1016/j.patter.2024.101046.

H. Feng, Y. Wang, Z. Li, N. Zhang, Y. Zhang, and Y. Gao, “Information Leakage in Deep Learning-Based Hyperspectral Image Classification: A Survey,” Remote Sens., vol. 15, no. 15, Art. no. 3793, 2023, doi: 10.3390/rs15153793.

G. Gallitto et al., “External validation of machine learning models—registered models and adaptive sample splitting,” GigaScience, vol. 14, Art. no. giaf036, 2025, doi: 10.1093/gigascience/giaf036.

S. S. Band et al., “Application of explainable artificial intelligence in medical health: A systematic review of interpretability methods,” Inform. Med. Unlocked, vol. 40, Art. no. 101286, 2023, doi: 10.1016/j.imu.2023.101286.

W. Chmiel, J. Kwiecień, and K. Motyka, “Saliency Map and Deep Learning in Binary Classification of Brain Tumours,” Sensors, vol. 23, no. 9, Art. no. 4543, 2023, doi: 10.3390/s23094543.

J. Sigut, F. Fumero, J. Estévez, S. Alayón, and T. Díaz-Alemán, “In-Depth Evaluation of Saliency Maps for Interpreting Convolutional Neural Network Decisions in the Diagnosis of Glaucoma Based on Fundus Imaging,” Sensors, vol. 24, no. 1, Art. no. 239, 2024, doi: 10.3390/s24010239.

Y. Gao, J. Liu, W. Li, M. Hou, Y. Li, and H. Zhao, “Augmented Grad-CAM++: Super-Resolution Saliency Maps for Visual Interpretation of Deep Neural Network,” Electronics, vol. 12, no. 23, Art. no. 4846, 2023, doi: 10.3390/electronics12234846.

K. Venkatesh, S. Mutasa, F. Moore, J. Sulam, and P. H. Yi, “Gradient-Based Saliency Maps Are Not Trustworthy Visual Explanations of Automated AI Musculoskeletal Diagnoses,” Journal of Imaging Informatics in Medicine, vol. 37, no. 5, pp. 2490–2499, 2024, doi: 10.1007/s10278-024-01136-4.

A. Baumann et al., “Neural illumination calibration for surgical workflow-optimized spectral imaging,” International Journal of Computer Assisted Radiology and Surgery, vol. 21, no. 4, pp. 665–675, 2026, doi: 10.1007/s11548-025-03525-8.

S. Nie et al., “Hyperspectral imaging combined with deep learning models for the prediction of geographical origin and fungal contamination in millet,” Front. Sustain. Food Syst., vol. 8, Art. no. 1454020, 2024, doi: 10.3389/fsufs.2024.1454020.