Page Header Logo Applied Science and Engineering Progress

Home > ONLINE FIRST > Klaiklueng

A comparative evaluation of agentic AI-assisted model and visual inspection for a deep learning analysis model: An empirical study of lime quality classification in compliance with Codex Standard 217-1999

Alongkorn Klaiklueng, Chavittha Kengpol

Abstract


Accurate quality classification is essential in fruit trade, particularly for peel-based grading. However, conventional visual inspection is often subjective and inconsistent. The literature review identifies a research gap: no Codex-aligned lime grading model has applied an agentic AI-assisted model while systematically comparing regression models. This gap highlights the need to align data structure, model assumptions, and predictive performance in the development of standard-aligned classification models. The objective of this research is to design a deep learning (DL)-based image analysis model incorporating an agentic AI-assisted model for fruit quality classification and to evaluate its performance against conventional visual inspection. Experimental results show that the proposed model achieves over 90% classification accuracy, significantly outperforming visual inspection (55%). The model integrates Visual Geometry Group 19 (VGG19) for the image classification model with ordinal logistic regression (OLR) to classify lime quality attributes. VGG19 reached 94% accuracy, while OLR showed significant predictors and strong ordinal associations, confirming its suitability for Codex-aligned quality grading. The model also provides explainable outputs through measured lime attributes and class probabilities. The contribution of this research is the development of a Codex-aligned lime quality classification model that integrates a Convolutional Neural Network (CNN)-based deep learning model, OLR, and a Large Language Model (LLM)-driven agentic AI-assisted model for automated and interpretable quality assessment based upon Codex Standard 217-1999. Its advantage lies in improving inspection standardization, reducing subjectivity, and enhancing grading efficiency. The benefit of the proposed model is improving quality control reliability while reducing missed produce sales opportunities for farmers.

Keywords



[1] M. Abou Ali, F. Dornaika, and J. Charafeddine, “Agentic AI: A comprehensive survey of architectures, applications, and future directions,” Artificial Intelligence Review, vol. 59, Art. no. 11, 2026, doi: 10.1007/s10462-025-11422-4.

[2] J. E. See, “Visual inspection reliability for precision manufactured parts,” Human Factors, vol. 57, no. 8, pp. 1427–1442, 2015, doi: 10.1177/0018720815602389.

[3] PictureThis, “Key lime citrus,” 2026. [Online]. Available: https://www.picturethisai.com/wiki/C itrus_aurantiifolia.html. Accessed: Jan. 26, 2026.

[4] World Integrated Trade Solution (WITS), “Lemons and limes, fresh or dried exports by country 2024,” 2024. [Online]. Available: https://wits. worldbank.org. Accessed: Jan. 14, 2026.

[5] N. Parashar, P. Johri, A. A. Khan, N. Gaur, and S. Kadry, “An integrated analysis of yield prediction models: A comprehensive review of advancements and challenges,” Computers, Materials & Continua, vol. 80, no. 1, pp. 389–425, 2024, doi: 10.32604/cmc.2024.050240.

[6] V. K. Vishnoi, K. Kumar, B. Kumar, S. Mohan, and A. A. Khan, “Detection of apple plant diseases using leaf images through convolutional neural network,” IEEE Access, vol. 11, pp. 6594-6609, 2023, doi: 10.1109/ACCESS.2022.3232917.

[7] A. A. Khan, M. Faheem, R. N. Bashir, C. Wechtaisong, and M. Z. Abbas, “Internet of Things (IoT) assisted context aware fertilizer recommendation,” IEEE Access, vol. 10, pp. 129505–129520, 2022, doi: 10.1109/ACCESS. 2022.3228160.

[8] A. A. Khan, M. A. Nauman, R. N. Bashir, R. Jahangir, R. Alroobaea, A. Binmahfoudh, M. Alsafyani, and C. Wechtaisong, “Context aware evapotranspiration (ETs) for saline soils reclamation,” IEEE Access, vol. 10, pp. 110050–110064, 2022, doi: 10.1109/A.CCESS.2022.3206009.

[9] R. N. Bashir, F. A. Khan, A. A. Khan, M. Tausif, M. Z. Abbas, M. M. A. Shahid, and N. Khan, “Intelligent optimization of reference evapotranspiration (ETo) for precision irrigation,” Journal of Computational Science, vol. 69, Art. no. 102025, 2023, doi: 10.1016/j.jocs. 2023.102025.

[10] S. C. Ramu, D. Raman, K. Ramana, A. V. Krishna, A. A. Khan, S. Khan, and S. M. Abdu, “A lightweight deep learning and whale optimization framework for sustainable precision agriculture,” Discover Computing, vol. 29, Art. no. 73, 2026, doi: 10.1007/s10791-026-09952-8.

[11] K. Vijaipriya, N. Moses, and P. A. Michael, “Efficient multi-task learning in multi-user multiple input multiple output systems integrated orthogonal frequency division multiplexing systems: A hybrid amalgamated convolutional neural network–bidirectional long short-term memory approach,” Applied Science and Engineering Progress, vol. 19, no. 2, Art. no. 7942, 2026, doi: 10.14416/j.asep.2025.11.005.

[12] S. R. Kumar and M. P. Chinnasamy, “AI-Driven Detection of Tomato Leaf Diseases for Sustainable Agriculture,” Applied Science and Engineering Progress, vol. 18, no. 4, Art. no. 7825, 2025, doi: 10.14416/j.asep.2025.06.007.

[13] D. B. Acharya, K. Kuppan, and B. Divya, “Agentic AI: Autonomous intelligence for complex goals—A comprehensive survey,” IEEE Access, vol. 13, pp. 18912–18936, 2025, doi: 10.1109/ACCESS.2025.3532853.

[14] J. Lee and H. Su, “Agentic AI for smart manufacturing,” Manufacturing Letters, vol. 46, pp. 92–96, 2025, doi: 10.1016/j.mfglet.2025.10.013.

[15] A. Loaiza-Bonilla, C. Yost, S. Kurnaz, E. Tuysuz, N. G. Thaker, D. Giritlioglu, and J. P. Noel Meza, “Transforming oncology clinical trial matching through neuro-symbolic, multi-agent AI and an oncology-specific knowledge graph: A prospective evaluation in 3804 patients,” ESMO Real World Data and Digital Oncology, vol. 12, Art. no. 100706, 2026, doi: 10.1016/j.esmorw. 2026.100706.

[16] FAS-USDA, “Production: Lemons & Limes.” [Online]. Available: https://www.fas.usda.gov/data/ production/commodity/0572120. Accessed: Feb. 18, 2026.

[17] Codex Alimentarius Commission, Standard for Mexican Limes (CXS 217-1999), Rome, Italy, 1999.

[18] A. M. Torkashvand, A. Ahmadi, and N. L. Nikravesh, “Prediction of kiwifruit firmness using fruit mineral nutrient concentration by artificial neural network (ANN) and multiple linear regressions (MLR),” Journal of Integrative Agriculture, vol. 16, no. 7, pp. 1634–1644, 2017, doi: 10.1016/S2095-3119(16)61546-0.

[19] X. Qu, L. Ma, T. Shen, X. Han, X. Yan, and S. Bi, “Ordinal regression-extreme learning machine based apple grading method,” in Proceedings of the 2022 China Automation Congress (CAC), 2022, pp. 3507–3511, doi: 10.1109/CAC57257.2022.10054740.

[20] Q. Dong, X. Chen, and B. Huang, Data Analysis in Pavement Engineering: Methodologies and Applications. Cambridge, U.K.: Woodhead Publishing, 2023.

[21] P. A. Gutiérrez, M. Pérez-Ortiz, J. Sánchez-Monedero, F. Fernández-Navarro, and C. Hervás-Martínez, “Ordinal regression methods: Survey and experimental study,” IEEE Transactions on Knowledge and Data Engineering, vol. 28, no. 1, pp. 127–146, 2016, doi: 10.1109/TKDE.2015.2457911.

[22] J. Cohen, “A coefficient of agreement for nominal scales,” Educational and Psychological Measurement, vol. 20, no. 1, pp. 37–46, 1960, doi: 10.1177/001316446002000104.

[23] M. L. McHugh, “Interrater reliability: The kappa statistic,” Biochemia Medica, vol. 22, no. 3, pp. 276–282, 2012, doi: 10.11613/BM.2012.031.

[24] O. Tasneem and R. Pieters, “Human–robot collaborative visual inspection with Large Language Models,” Robotics and Computer-Integrated Manufacturing, vol. 98, Art. no. 103154, 2026, doi: 10.1016/j.rcim.2025.103154.

[25] A. Corallo, V. Del Vecchio, A. Di Prizio, and M. Buscicchio, “Benchmarking AI-based visual inspection systems for aerospace quality control: A multi-vendor comparative evaluation,” Procedia Computer Science, vol. 277, pp. 1621–1630, 2026, doi: 10.1016/j.procs.2026.02.200.

[26] L. Ierimonti, F. Mariani, F. Ubertini, and I. Venanzi, “A dynamic Bayesian networks-based approach for data fusion of model-based SHM and visual inspections in post-tensioned bridges,” Structures, vol. 86, Art. no. 111389, 2026, doi: 10.1016/j.istruc.2026.111389.

[27] M. A. Saputra and I. Nurhaida, “Signature originality verification using a deep learning approach,” Electronic Journal of Education, Social Economics and Technology, vol. 5, no. 1, pp. 19–29, 2024, doi: 10.33122/ejeset.v5i1.310.

[28] C. Shorten and T. M. Khoshgoftaar, “A survey on image data augmentation for deep learning,” Journal of Big Data, vol. 6, no. 1, Art. no. 60, 2019, doi: 10.1186/s40537-019-0197-0.

[29] R. R. Selvaraju, M. Cogswell, A. Das, R. Vedantam, D. Parikh, and D. Batra, “Grad-CAM: Visual explanations from deep networks via gradient-based localization,” in Proceedings of the 2017 IEEE International Conference on Computer Vision (ICCV), Venice, Italy, 2017, pp. 618–626, doi: 10.1109/ICCV.2017.74.

Full Text: PDF

DOI: 10.14416/j.asep.2026.08.014

Refbacks

  • There are currently no refbacks.


Copyright © 2025 by King Mongkut’s University of Technology North Bangkok. All rights reserved.