Citation :

Aljon L. Abines, Aimee D. Molato, "Evaluating Denoising Models for Feature Enhancement and Improved SVM-Based Classification of Locally Made Earthen Ceramic Pots," International Journal of Engineering Trends and Technology (IJETT), vol. 73, no. 12, pp. 194-206, 2025. Crossref, https://doi.org/10.14445/22315381/IJETT-V73I12P116

Abstract

Manual inspection of locally made earthen ceramic pots suffers from inconsistency and subjectivity, creating problems for quality control in traditional pottery production. This research examines how denoising affects feature extraction in SVM-based ceramic pot classification. The study compares three deep learning denoising architectures: the Denoising Autoencoder with Convolutional Autoencoder (DAE-CAE), Denoising Convolutional Neural Network (DnCNN), and a Generative Adversarial Network (GAN). PSNR, SSIM, and RMSE as metrics were used for performance evaluation. Results show that the DAE-CAE Model outperforms the other architectures, achieving a PSNR of 23.2087, SSIM of 0.4828, and RMSE of 0.0713, while DnCNN reaches 23.0786 PSNR, 0.4742 SSIM, and 0.0725 RMSE, and GAN achieves 23.1815 PSNR, 0.4784 SSIM, and 0.0719 RMSE. Features extracted from DAE-CAE-denoised images are used to train classifiers, and SVM achieves 93.23% accuracy. This exceeds both Random Forest at 90.73% and CNN at 90.20%. The results indicate that denoising improves classification generalizability, precision, and reliability. The DAE-CAE-enhanced SVM framework proves most effective for this task. Combining deep learning denoising with SVM provides a practical automated alternative to manual inspection, offering both improved accuracy and potential for scaling quality assessment in traditional ceramic production.

Keywords

Deep Learning, Earthen ceramic pot classification, Feature enhancement, Image denoising models, Support Vector Machine.

References

[1] Khaled Alomar, Halil Ibrahim Aysel, and Xiaohao Cai, “Data Augmentation in Classification and Segmentation: A Survey and New Strategies,” Journal of Imaging, vol. 9, no. 2, pp. 1-26, 2023.
[CrossRef] [Google Scholar] [Publisher Link]

[2] Markus Bayer, Marc-André Kaufhold, and Christian Reuter, “A Survey on Data Augmentation for Text Classification,” ACM Computing Surveys, vol. 55, no. 7, pp. 1-39, 2022.
[CrossRef] [Google Scholar] [Publisher Link]

[3] Tom Brown et al., “Language Models are Few-Shot Learners,” 34^th Conference on Neural Information Processing Systems, Vancouver, Canada, pp. 1877-1901, 2020.
[Google Scholar] [Publisher Link]

[4] Jair Cervantes et al., “A Comprehensive Survey on Support Vector Machine Classification: Applications, Challenges, and Trends,” Neurocomputing, vol. 408, pp. 189-215, 2020.
[CrossRef] [Google Scholar] [Publisher Link]

[5] Huseyin Coskun, Tuncay Yi̇ği̇t, and İsmail Serkan Üncü, “Integration of Digital Quality Control for Intelligent Manufacturing of Industrial Ceramic Tiles,” Ceramics International, vol. 48, no. 23, pp. 34210-34233, 2022.
[CrossRef] [Google Scholar] [Publisher Link]

[6] Esteban Cumbajin et al., “A Real-Time Automated Defect Detection System for Ceramic Pieces Manufacturing Process based on Computer Vision with Deep Learning,” Sensors, vol. 24, no. 1, pp. 1-22, 2023.
[CrossRef] [Google Scholar] [Publisher Link]

[7] Steffen Dereich, and Arnulf Jentzen, “Convergence Rates for the Adam Optimizer,” arXiv Preprint, pp. 1-43, 2024.
[CrossRef] [Google Scholar] [Publisher Link]

[8] Timothy O. Hodson, “Root Mean Square Error (RMSE) or Mean Absolute Error (MAE): When to Use them or Not,” Geoscientific Model Development Discussions, vol. 15, no. 14, pp. 1-10, 2022.
[CrossRef] [Google Scholar] [Publisher Link]

[9] Shafaf Ibrahim et al., “Automated Defective Ceramic Tiles Classification using Image Processing Techniques,” Journal of Advanced Research in Applied Sciences and Engineering Technology, vol. 32, no. 3, pp. 355-365, 2023.
[CrossRef] [Google Scholar] [Publisher Link]

[10] Imran, Naeem Iqbal, and Do-Hyeun Kim, “Intelligent Material Data Preparation Mechanism based on Ensemble Learning for AI-Based Ceramic Material Analysis,” Advanced Theory and Simulations, vol. 5, no. 11, 2022.
[CrossRef] [Google Scholar] [Publisher Link]

[11] Vijay Janapa Reddi et al., “MLPerf Mobile Inference Benchmark: An Industry-Standard Open-Source Machine Learning Benchmark for On-Device AI,” Proceedings of the 5^th MLSys Conference, Santa Clara, CA, USA, pp. 352-369, 2022.
[Google Scholar] [Publisher Link]

[12] Taehyeon Kim et al., “Comparing Kullback-Leibler Divergence and Mean Squared Error Loss in Knowledge Distillation,” arXiv Preprint, pp. 1-11, 2021.
[CrossRef] [Google Scholar] [Publisher Link]

[13] See Alex Krizhevsky, Ilya Sutskever, and Geoffrey E. Hinton, From Photographic Image to Computer Vision, Monitoring Laws Profiling and Identity in the World State, Cambridge University Press, pp. 135-157, 2019.
[CrossRef] [Google Scholar] [Publisher Link]

[14] Yuh-Jye Lee, and O.L. Mangasarian, “SSVM: A Smooth Support Vector Machine for Classification,” Computational Optimization and Applications, vol. 20, no. 1, pp. 5-22, 2001.
[CrossRef] [Google Scholar] [Publisher Link]

[15] Jingyun Liang et al., “Swinir: Image Restoration using Swin Transformer,” Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) Workshops, pp. 1833-1844, 2021.
[Google Scholar] [Publisher Link]

[16] Jakub Nalepa, and Michal Kawulok, “Selecting Training Sets for Support Vector Machines: A Review,” Artificial Intelligence Review, vol. 52, no. 2, pp. 857- 900, 2018.
[CrossRef] [Google Scholar] [Publisher Link]

[17] K. Nanthini et al., “A Survey on Data Augmentation Techniques,” 2023 7^th International Conference on Computing Methodologies and Communication (ICCMC), Erode, India, pp. 913-920, 2023.
[CrossRef] [Google Scholar] [Publisher Link]

[18] Jim Nilsson, and Tomas Akenine-Möller, “Understanding SSIM,” arXiv Preprint, pp. 1-8, 2020.
[CrossRef] [Google Scholar] [Publisher Link]

[19] Simone Parisotto et al., “Unsupervised Clustering of Roman Potsherds via Variational Autoencoders,” arXiv Preprint, pp. 1-16, 2022.
[CrossRef] [Google Scholar] [Publisher Link]

[20] Derek A. Pisner, and David M. Schnyer, Support Vector Machine, Machine Learning: Methods and Applications to Brain Disorders, pp. 101-121, 2020.
[CrossRef] [Google Scholar] [Publisher Link]

[21] Edisson Pugo-Mendez, and Luis Serpa-Andrade, “Development of a Platform based on Artificial Vision with SVM and KNN Algorithms for the Identification and Classification of Ceramic Tiles,” Artificial Intelligence and Social Computing, vol. 28, no. 28, pp. 173-181, 2022.
[CrossRef] [Google Scholar] [Publisher Link]

[22] Yuriy Reznik, “Another Look at SSIM Image Quality Metric,” Electronic Imaging, vol. 35, pp. 1-7, 2023.
[CrossRef] [Google Scholar] [Publisher Link]

[23] Roy Schwartz et al., “Green AI,” Communications of the ACM, vol. 63, no. 12, pp. 54-63, 2020.
[CrossRef] [Google Scholar] [Publisher Link]

[24] De Rosal Igantius Moses Setiadi, “PSNR vs SSIM: Imperceptibility Quality Assessment for Image Steganography,” Multimedia Tools and Applications, vol. 80, no. 6, pp. 8423-8444, 2020.
[CrossRef] [Google Scholar] [Publisher Link]

[25] Connor Shorten, and Taghi M. Khoshgoftaar, “A Survey on Image Data Augmentation for Deep Learning,” Journal of Big Data, vol. 6, no. 1, pp. 1-48, 2019.
[CrossRef] [Google Scholar] [Publisher Link]

[26] Okeke Stephen, Uchenna Joseph Maduh, and Mangal Sain, “A Machine Learning Method for Detection of Surface Defects on Ceramic Tiles Using Convolutional Neural Networks,” Electronics, vol. 11, no, 1, pp. 1-22, 2021.
[CrossRef] [Google Scholar] [Publisher Link]

[27] Chunwei Tian  et al., “Deep Learning on Image Denoising: An Overview,” Neural Networks, vol. 131, pp. 251-275, 2020.
[CrossRef] [Google Scholar] [Publisher Link]

[28] Pascal Vincent et al., “Stacked Denoising Autoencoders: Learning Useful Representations in a Deep Network with a Local Denoising Criterion,” Journal of Machine Learning Research, vol. 11, no. 12, pp. 3371-3408, 2010.
[Google Scholar] [Publisher Link]

[29] Slamet Widodo, Herlambang Brawijaya, and Samudi Samudi, “Stratified K-Fold Cross Validation Optimization on Machine Learning for Prediction,” Sinkron: Jurnal Dan Penelitian Teknik Informatika, vol. 6, no. 4, pp. 2407-2414, 2022.
[CrossRef] [Google Scholar] [Publisher Link]

[30] Min-Ling Zhu, Liang-Liang Zhao, and Li Xiao, “Image Denoising based on GAN with Optimization Algorithm,” Electronics, vol. 11, no. 15, pp. 1-12, 2022.
[CrossRef] [Google Scholar] [Publisher Link]

[31] Bo Zhang et al., “Signal Data Augmentation Algorithm Research based on Fractal Theory,” 2024 IEEE 7^th International Conference on Information Systems and Computer Aided Education (ICISCAE), Dalian, China, pp. 1023-1026, 2024.
[CrossRef] [Google Scholar] [Publisher Link]

[32] Houwang Zhang, Yuan Zhu, and Hanying Zheng, “NAMF: A Non-Local Adaptive mean Filter for Salt-and-Pepper Noise Removal,” arXiv Preprint, pp. 1-9, 2020.
[CrossRef] [Google Scholar] [Publisher Link]