Fruit Ripeness Analysis Using Colour Transition Matrix and Eigenvalues
Journal of Applied Informatics and Computing · 16.04.2026 · 10 (2) · 1404-1415
DOI: 10.30871/jaic.v10i2.12382 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Applied Informatics and Computing, online erschienen am 16.04.2026. Verfasst von Septiana Sulastri Septi und 1 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY-SA 4.0).
Abstract (Originalsprache)
Fruit ripeness classification plays an important role in maintaining fruit quality, reducing post-harvest losses, and ensuring consistent grading in agricultural supply chains. Bananas, particularly the Cavendish variety, are widely traded globally and require reliable methods for determining ripeness levels during distribution and storage. However, conventional visual inspection methods remain subjective and often lead to inconsistent evaluations. This study proposes an interpretable mathematical approach for fruit ripeness analysis based on the Colour Dominant–Transition Matrix (CDTM) combined with eigenvalue analysis. The dataset consists of 3,495 real images of Cavendish bananas obtained from an open-access dataset. Image preprocessing includes object segmentation, colour normalization, and noise reduction to ensure illumination consistency. Dominant colour features are extracted using K-Means clustering in the CIELAB colour space, and the CDTM is constructed to model spatial colour transition probabilities. The second-largest eigenvalue (λ₂) is used as a quantitative indicator of colour stability and homogeneity. Experimental results show that optimally ripe bananas produce the lowest λ₂ value (0.1045), indicating high colour uniformity, while unripe bananas exhibit the highest λ₂ value (0.7972). Using λ₂ as a single feature, an SVM classifier achieved an accuracy of 64%. Although the classification performance remains moderate, the proposed CDTM-based eigenvalue feature provides an interpretable and computationally efficient indicator suitable for non-destructive fruit ripeness monitoring systems.
Zitieren
APA
Septi, S. S., & Azhar, Y. (2026). Fruit Ripeness Analysis Using Colour Transition Matrix and Eigenvalues. Journal of Applied Informatics and Computing, 10(2), 1404-1415. https://doi.org/10.30871/jaic.v10i2.12382
Vancouver
Septi SS, Azhar Y. Fruit Ripeness Analysis Using Colour Transition Matrix and Eigenvalues. Journal of Applied Informatics and Computing. 2026;10(2):1404-1415. doi:10.30871/jaic.v10i2.12382
BibTeX
@article{Septi2026,
author = {Septi, Septiana Sulastri and Azhar, Yufis},
title = {Fruit Ripeness Analysis Using Colour Transition Matrix and Eigenvalues},
journal = {Journal of Applied Informatics and Computing},
year = {2026},
volume = {10},
number = {2},
pages = {1404--1415},
doi = {10.30871/jaic.v10i2.12382}
}
EUVIMED