Hybrid AI Pipeline for Laboratory Detection of Internal Potato Defects Using 2D RGB Imaging

Hamdi, Slim; Loukil, Kais; Boubaker, Adem Haj; Snoussi, Hichem; Abid, Mohamed

doi:10.3390/jimaging11120431

This is an early access version, the complete PDF, HTML, and XML versions will be available soon.

Open AccessArticle

Hybrid AI Pipeline for Laboratory Detection of Internal Potato Defects Using 2D RGB Imaging

by

Slim Hamdi

^1,2,3,*

,

Kais Loukil

¹

,

Adem Haj Boubaker

³

,

Hichem Snoussi

^2,4 and

Mohamed Abid

^1,5

¹

CES Laboratory, ENIS National Engineering School, University of Sfax, Sfax B.P. 3038, Tunisia

²

LIST3N, Université de Technologie de Troyes, 10300 Troyes, France

³

Eurocelp SAS, 14 Rue des Prés de Lyon, La Chapelle Saint Luc, Troyes 10600, France

⁴

EUT+ Data Science Institute, European Union

⁵

Lab-STICC/ENSIBS, University of Bretagne du Sud, 56100 Lorient, France

^*

Author to whom correspondence should be addressed.

J. Imaging 2025, 11(12), 431; https://doi.org/10.3390/jimaging11120431

Submission received: 19 September 2025 / Revised: 12 November 2025 / Accepted: 15 November 2025 / Published: 3 December 2025

(This article belongs to the Special Issue Imaging Applications in Agriculture)

Download Versions Notes

Abstract

The internal quality assessment of potato tubers is a crucial task in agro-laboratory processing. Traditional methods struggle to detect internal defects such as hollow heart, internal bruises, and insect galleries using only surface features. We present a novel, fully modular hybrid AI architecture designed for defect detection using RGB images of potato slices, suitable for integration in laboratory. Our pipeline combines high-recall multi-threshold YOLO detection, contextual patch validation using ResNet, precise segmentation via the Segment Anything Model (SAM), and skin-contact analysis using VGG16 with a Random Forest classifier. Experimental results on a labeled dataset of over 6000 annotated instances show a recall above 95% and precision near 97.2% for most defect classes. The approach offers both robustness and interpretability, outperforming previous methods that rely on costly hyperspectral or MRI techniques. This system is scalable, explainable, and compatible with existing 2D imaging hardware.

Keywords: potato defect detection; deep learning; RGB imaging; YOLO; SAM; ResNet; random forest

Share and Cite

MDPI and ACS Style

Hamdi, S.; Loukil, K.; Boubaker, A.H.; Snoussi, H.; Abid, M. Hybrid AI Pipeline for Laboratory Detection of Internal Potato Defects Using 2D RGB Imaging. J. Imaging 2025, 11, 431. https://doi.org/10.3390/jimaging11120431

AMA Style

Hamdi S, Loukil K, Boubaker AH, Snoussi H, Abid M. Hybrid AI Pipeline for Laboratory Detection of Internal Potato Defects Using 2D RGB Imaging. Journal of Imaging. 2025; 11(12):431. https://doi.org/10.3390/jimaging11120431

Chicago/Turabian Style

Hamdi, Slim, Kais Loukil, Adem Haj Boubaker, Hichem Snoussi, and Mohamed Abid. 2025. "Hybrid AI Pipeline for Laboratory Detection of Internal Potato Defects Using 2D RGB Imaging" Journal of Imaging 11, no. 12: 431. https://doi.org/10.3390/jimaging11120431

APA Style

Hamdi, S., Loukil, K., Boubaker, A. H., Snoussi, H., & Abid, M. (2025). Hybrid AI Pipeline for Laboratory Detection of Internal Potato Defects Using 2D RGB Imaging. Journal of Imaging, 11(12), 431. https://doi.org/10.3390/jimaging11120431

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Article metric data becomes available approximately 24 hours after publication online.

Article Menu

Hybrid AI Pipeline for Laboratory Detection of Internal Potato Defects Using 2D RGB Imaging

Abstract

Share and Cite

Article Metrics

Further Information

Guidelines

MDPI Initiatives

Follow MDPI